modules/learningmethods.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyPVe/2blsN8j2J8C64BE8bi",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ekrell/ams_ai_shortcourse_2024/blob/main/modules/learningmethods.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **AMS 2024 AI Short Course: Learning Methods in AI**\n",
        "\n",
        "---\n",
        "\n",
        "*Notebook by Maria J. Molina (University of Maryland, College Park)*"
      ],
      "metadata": {
        "id": "pwgWAFL8GZWw"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "This Colab Notebook demonstrates different types of learning methods in artificial intelligence, with a specific focus on supervised and unsupervised learning, along with regression and classification problems."
      ],
      "metadata": {
        "id": "-Eo_OAT7Ppqn"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## **Supervised Learning**\n",
        "\n",
        "---\n",
        "\n",
        "Supervised learning refers to machine learning tasks where the dataset used contains both the predictors (x) and predictands (y) *a priori.* In other words, these are situations where we know the \"answer(s)\" beforehand. This information is useful for supervised tasks because the machine learning model can receive information about the skill of its own predictions based on the known predictands, and then these signals are used to update model weights accordingly to improve subsequent predictions. The predictand(s) can be continuous or discrete, which would lend itself to a regression or classification type of problem respectively.\n",
        "\n",
        "We often refer to the predictands as \"labels,\" particularly in the case of classification problems.\n",
        "\n"
      ],
      "metadata": {
        "id": "cHloNSeDZ2eU"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![supervised](https://drive.google.com/uc?export=view&id=1g74XtOQO7jZsEUxHEDnD4aKYAOOnhQjE)"
      ],
      "metadata": {
        "id": "leLgZajv6xGk"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Labels can be of different data types, such as:\n",
        "\n",
        "*   Single point value\n",
        "*   Single image pixel (i.e., grid-cell)\n",
        "*   Image-based (i.e., spatial in nature)\n",
        "*   Temporal (i.e., time series)\n",
        "\n",
        "In supervised learning, we can have regression or classification problems, which are further detailed below."
      ],
      "metadata": {
        "id": "3Nh2ooRNFj7W"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Regression\n",
        "\n",
        "---\n",
        "\n",
        "**In regression problems**, we aim to predict the specific value of *something*. These values are generally continuous in nature. A simple example from statistics would be a linear regression y = mx + b, where y (our predictand) would be a continuous value that we can predict using our input x (our predictor)."
      ],
      "metadata": {
        "id": "esxC_D-RPp0z"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![supervised](https://drive.google.com/uc?export=view&id=1itJDHx1gq4WqEQ00hxKafMQ9-A9BjEhB)\n"
      ],
      "metadata": {
        "id": "ntxfSbLh-ADd"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Some examples of regression problems in the environmental sciences could include:\n",
        "\n",
        "*   tropical cyclone maximum wind speed at a certain lead time\n",
        "*   maximum land surface temperature in an urban area\n",
        "*   percentage of cloud coverage over a predifined domain\n",
        "*   bias correction of subseasonal forecasts of precipitation\n",
        "\n",
        "Importantly, when using various input variables (i.e., features or channels), normalizing or standardizing the variables is needed to make sure they have a similar dynamic range. Otherwise, some input variables may have much larger magnitudes than other input variables, limiting their contributions to the learning process. By normalizing/standardizing our features, we facilitate meaningful comparisons between features. This practice can also improve machine learning model convergence.\n",
        "\n",
        "Standardization involves centering the respective data to have a mean of zero and a standard deviation of one, which can be achieved by subtracting the mean and dividing by the standard deviation (i.e., z-score). Normalization involves the scaling of data to a set range, such as from 0 to +1 or -1 to +1, by using the minimum and maximum values of the respective variable."
      ],
      "metadata": {
        "id": "d-uDePYn-apy"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Here we have some helper functions for normalizing and standardizing our data using numpy."
      ],
      "metadata": {
        "id": "WmqM9d_sEJHu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "\n",
        "def z_score(da):\n",
        "    \"\"\"\n",
        "    Compute z-score for training data.\n",
        "    Normalizing data for ML model training.\n",
        "    Args:\n",
        "        da (array): Array of training variable.\n",
        "    \"\"\"\n",
        "    return (da - np.mean(da)) / np.std(da)\n",
        "\n",
        "def min_max_scale(da):\n",
        "    \"\"\"\n",
        "    Scale training data by minimum and maximum values.\n",
        "    Args:\n",
        "        da (array): Array of training variable.\n",
        "    \"\"\"\n",
        "    return (da - np.min(da)) / (np.max(da) - np.min(da))"
      ],
      "metadata": {
        "id": "uA1hA1WUC-R-"
      },
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "We can create a little random array just to visualize what the normalization and standardization methods do to our data."
      ],
      "metadata": {
        "id": "FMwM6y2KFe0J"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "np.random.seed(0)\n",
        "sample_array = np.random.rand(10)\n",
        "print(sample_array)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "J44BpbaEEi3g",
        "outputId": "6bac4730-2e50-419e-cb36-36e6b7afeddf"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[0.5488135  0.71518937 0.60276338 0.54488318 0.4236548  0.64589411\n",
            " 0.43758721 0.891773   0.96366276 0.38344152]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.hist(sample_array)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "4J-FyS5iE13t",
        "outputId": "c371fd4b-e78d-46c9-ea91-f55e162d8490"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.hist(z_score(sample_array))\n",
        "plt.title('z-score')\n",
        "plt.show()\n",
        "\n",
        "plt.hist(min_max_scale(sample_array))\n",
        "plt.title('min-max')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 887
        },
        "id": "GJ2xJLedE2K6",
        "outputId": "eea22718-9ad7-454f-c54d-90ba65e7024a"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAigAAAGzCAYAAAAFROyYAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAl80lEQVR4nO3df1TUdaL/8ReoDLoJxFV+qAgkhZoGpqnQXcFdily2I/d0y2W7i7nqVitndenmlc7ePOTZQ2Wm3Xst63aMrb2urZV6Vk0uYeqaaKFwUis3XQVrGawtGaFCZd7fP/br7J1g0EEG3sLzcc7nj/nM+z3zno8TPPvwGQgyxhgBAABYJLinFwAAAPBtBAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAeg1jjL7++uueXgaALkCgAPBy8uRJBQUF+dw6UlVVpezsbA0ZMkQDBw5UYmKifvrTn3qNcbvdeuaZZzR+/HiFhoZq6NChuuOOO1RVVeUZc+HCBS1btkyjRo2Sw+FQQkKCHnnkEbW0tHg9VkJCgn74wx+qrKxMkyZN0sCBA/X8889Lks6cOaNFixYpLi5ODodDSUlJeuKJJ+R2u7voSAEIpCBjjOnpRQCwR3NzszZu3Oi17/z58/rlL3+pkJAQnT59ut15p0+f1ujRozV06FDNnz9fEREROnnypN544w198MEHnnFz5sxRaWmpZsyYoezsbF24cEF//OMflZWVpYKCAknSfffdp9/85jf653/+Z02fPl379+/Xyy+/rNzcXK+1JSQkaMCAAfrrX/+q+++/XwkJCUpOTtbkyZOVlpamTz/9VPfff79GjhypvXv36pVXXtEvfvELrVq1qusPHICuZQDgEn7+85+bfv36mR07dvgcs3HjRiPJvPfeez7H7Nixw0gyv/jFL9rc53a7jTHG1NTUGElm3rx5Xvf/67/+q5HktYb4+HgjyWzfvt1r7LJly8x3vvMd86c//clr/5IlS0y/fv1MXV2d7xcLwAr8iAdAh15++WU9++yzevLJJzV9+nSf4yIiIiRJW7Zs0fnz59sd8/rrrysoKEhLly5tc9/FHx9t27ZNklRYWOh1/0MPPSRJ2rp1q9f+xMREZWdne+3bsGGDvvvd7+raa6/V559/7tmysrLU2tqq3bt3d/CKAdigf08vAIC9ampq9MADDygvL88TDE1NTWpqavKM6devn4YOHaqMjAzdddddKi4u1sqVK5WZmanc3Fz9+Mc/lsPhkCQdP35cw4YNU2RkpM/nrK2tVXBwsJKSkrz2x8TEKCIiQrW1tV77ExMT2zzGxx9/rPfff19Dhw5t9zl8/ZgKgD0IFADt+vLLL3XXXXfphhtu0IsvvujZ/9RTT6m4uNhzOz4+3nNh7WuvvaZ9+/bpD3/4g8rKyvTTn/5UK1as0L59+3TNNdf49fyXuiD3ooEDB7bZ53a7ddttt2nx4sXtzrnhhhv8WguA7kegAGjD7Xbr3nvv1ZkzZ/TWW29p0KBBnvvy8/P1j//4j57b3w6EqVOnaurUqfr1r3+tdevW6d5779X69es1b948jRo1SmVlZfriiy98nkWJj4+X2+3Wxx9/rDFjxnj2NzQ06MyZM4qPj7/k+keNGqWmpiZlZWX5+9IBWIJrUAC0UVxcrLKyMv3ud79r8yOU6667TllZWZ7t1ltvlfS3My7mWx8KTE1NlSTPx4PvuusuGWO8zsBcdHHuD37wA0lq80mbp59+WpKUk5NzyfXfc889qqysVFlZWZv7zpw5owsXLlzyMQD0LD5mDMDLoUOHlJKSomnTpmnevHlt7v+Xf/mXduetWrVKzz77rP7pn/5Jo0aN0tmzZ/Xf//3fqq+vV01NjSd08vPz9corr2jGjBm644475Ha79cc//lHTp09v8zHje+65RxkZGXr33Xf1m9/8pt2PGY8bN05btmzxWstXX32l7373u3r//fd13333aeLEiWpubtahQ4f02muv6eTJkxoyZEhXHTIAgdCjnyECYJ23337bSPK5+XLw4EGTl5dnRo4caRwOh4mKijI//OEPTVVVlde4CxcumOXLl5vRo0ebkJAQM3ToUDNjxgxz4MABz5jz58+b4uJik5iYaAYMGGDi4uJMUVGR+eabb7weKz4+3uTk5LS7nrNnz5qioiKTlJRkQkJCzJAhQ0x6erp56qmnzLlz567gCAHoDpxBAQAA1uEaFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABY56r4Vfdut1t/+ctfNHjw4Mv++xwAAKBnGWN09uxZDRs2TMHB/p0TuSoC5S9/+Yvi4uJ6ehkAAKATTp06pREjRvg156oIlMGDB0v62wsMCwvr4dUAAIDL4XK5FBcX5/k+7o+rIlAu/lgnLCyMQAEA4CrTmcszuEgWAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1vErUJ577jnddNNNnl85n5aWpjfffLPDORs2bNDo0aMVGhqq8ePHa9u2bVe0YAAA0Pv5FSgjRozQ448/rgMHDqiqqkrf+973NHPmTB05cqTd8Xv37lVeXp7mzp2r6upq5ebmKjc3V4cPH+6SxQMAgN4pyBhjruQBIiMjtXz5cs2dO7fNfbNmzVJzc7O2bNni2Td16lSlpqZqzZo1l/0cLpdL4eHhamxs5I8FAgBwlbiS79+dvgaltbVV69evV3Nzs9LS0todU1lZqaysLK992dnZqqys7PCxW1pa5HK5vDYAANB39Pd3wqFDh5SWlqZvvvlG11xzjTZu3KixY8e2O9bpdCo6OtprX3R0tJxOZ4fPUVJSouLiYn+X1ikJS7Z2y/N0pZOP5/T0EgAACCi/z6AkJyerpqZG+/fv14MPPqjZs2frgw8+6NJFFRUVqbGx0bOdOnWqSx8fAADYze8zKCEhIUpKSpIkTZw4Ue+9956eeeYZPf/8823GxsTEqKGhwWtfQ0ODYmJiOnwOh8Mhh8Ph79IAAEAvccW/B8XtdqulpaXd+9LS0lRRUeG1r7y83Oc1KwAAAJKfZ1CKioo0Y8YMjRw5UmfPntW6deu0c+dOlZWVSZLy8/M1fPhwlZSUSJIWLlyojIwMrVixQjk5OVq/fr2qqqr0wgsvdP0rAQAAvYZfgXL69Gnl5+ervr5e4eHhuummm1RWVqbbbrtNklRXV6fg4L+flElPT9e6dev0q1/9So888oiuv/56bdq0SePGjevaVwEAAHqVK/49KN0hkL8HhU/xAAAQGD3ye1AAAAAChUABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHb8CpaSkRLfccosGDx6sqKgo5ebm6ujRox3OKS0tVVBQkNcWGhp6RYsGAAC9m1+BsmvXLi1YsED79u1TeXm5zp8/r9tvv13Nzc0dzgsLC1N9fb1nq62tvaJFAwCA3q2/P4O3b9/udbu0tFRRUVE6cOCApk2b5nNeUFCQYmJiOrdCAADQ51zRNSiNjY2SpMjIyA7HNTU1KT4+XnFxcZo5c6aOHDnS4fiWlha5XC6vDQAA9B2dDhS3261Fixbp1ltv1bhx43yOS05O1tq1a7V582b99re/ldvtVnp6uj755BOfc0pKShQeHu7Z4uLiOrtMAABwFQoyxpjOTHzwwQf15ptvas+ePRoxYsRlzzt//rzGjBmjvLw8LVu2rN0xLS0tamlp8dx2uVyKi4tTY2OjwsLCOrNcnxKWbO3Sx+sOJx/P6eklAABwSS6XS+Hh4Z36/u3XNSgXFRQUaMuWLdq9e7dfcSJJAwYM0IQJE3Ts2DGfYxwOhxwOR2eWBgAAegG/fsRjjFFBQYE2btyoHTt2KDEx0e8nbG1t1aFDhxQbG+v3XAAA0Df4dQZlwYIFWrdunTZv3qzBgwfL6XRKksLDwzVw4EBJUn5+voYPH66SkhJJ0mOPPaapU6cqKSlJZ86c0fLly1VbW6t58+Z18UsBAAC9hV+B8txzz0mSMjMzvfa/9NJLuu+++yRJdXV1Cg7++4mZL7/8UvPnz5fT6dS1116riRMnau/evRo7duyVrRwAAPRanb5ItjtdyUU2l8JFsgAABMaVfP/mb/EAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACs41eglJSU6JZbbtHgwYMVFRWl3NxcHT169JLzNmzYoNGjRys0NFTjx4/Xtm3bOr1gAADQ+/kVKLt27dKCBQu0b98+lZeX6/z587r99tvV3Nzsc87evXuVl5enuXPnqrq6Wrm5ucrNzdXhw4evePEAAKB3CjLGmM5O/uyzzxQVFaVdu3Zp2rRp7Y6ZNWuWmpubtWXLFs++qVOnKjU1VWvWrLms53G5XAoPD1djY6PCwsI6u9x2JSzZ2qWP1x1OPp7T00sAAOCSruT79xVdg9LY2ChJioyM9DmmsrJSWVlZXvuys7NVWVnpc05LS4tcLpfXBgAA+o7+nZ3odru1aNEi3XrrrRo3bpzPcU6nU9HR0V77oqOj5XQ6fc4pKSlRcXFxZ5cGC12NZ6okzlYBQE/p9BmUBQsW6PDhw1q/fn1XrkeSVFRUpMbGRs926tSpLn8OAABgr06dQSkoKNCWLVu0e/dujRgxosOxMTExamho8NrX0NCgmJgYn3McDoccDkdnlgYAAHoBv86gGGNUUFCgjRs3aseOHUpMTLzknLS0NFVUVHjtKy8vV1pamn8rBQAAfYZfZ1AWLFigdevWafPmzRo8eLDnOpLw8HANHDhQkpSfn6/hw4erpKREkrRw4UJlZGRoxYoVysnJ0fr161VVVaUXXnihi18KAADoLfw6g/Lcc8+psbFRmZmZio2N9WyvvvqqZ0xdXZ3q6+s9t9PT07Vu3Tq98MILSklJ0WuvvaZNmzZ1eGEtAADo2/w6g3I5vzJl586dbfbdfffduvvuu/15KgAA0Ifxt3gAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYx+9A2b17t+68804NGzZMQUFB2rRpU4fjd+7cqaCgoDab0+ns7JoBAEAv53egNDc3KyUlRatXr/Zr3tGjR1VfX+/ZoqKi/H1qAADQR/T3d8KMGTM0Y8YMv58oKipKERERfs8DAAB9T7ddg5KamqrY2Fjddttteueddzoc29LSIpfL5bUBAIC+I+CBEhsbqzVr1uj111/X66+/rri4OGVmZurgwYM+55SUlCg8PNyzxcXFBXqZAADAIn7/iMdfycnJSk5O9txOT0/X8ePHtXLlSr3yyivtzikqKlJhYaHntsvlIlIAAOhDAh4o7Zk8ebL27Nnj836HwyGHw9GNKwIAADbpkd+DUlNTo9jY2J54agAAcBXw+wxKU1OTjh075rl94sQJ1dTUKDIyUiNHjlRRUZE+/fRTvfzyy5KkVatWKTExUTfeeKO++eYbvfjii9qxY4f+93//t+teBQAA6FX8DpSqqipNnz7dc/vitSKzZ89WaWmp6uvrVVdX57n/3Llzeuihh/Tpp59q0KBBuummm/TWW295PQYAAMD/5XegZGZmyhjj8/7S0lKv24sXL9bixYv9XhgAAOi7+Fs8AADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6/gdKLt379add96pYcOGKSgoSJs2bbrknJ07d+rmm2+Ww+FQUlKSSktLO7FUAADQV/gdKM3NzUpJSdHq1asva/yJEyeUk5Oj6dOnq6amRosWLdK8efNUVlbm92IBAEDf0N/fCTNmzNCMGTMue/yaNWuUmJioFStWSJLGjBmjPXv2aOXKlcrOzvb36QEAQB8Q8GtQKisrlZWV5bUvOztblZWVPue0tLTI5XJ5bQAAoO/w+wyKv5xOp6Kjo732RUdHy+Vy6euvv9bAgQPbzCkpKVFxcXGglwZcUsKSrT29BFjq5OM5Pb2EPoH/BruHje9nKz/FU1RUpMbGRs926tSpnl4SAADoRgE/gxITE6OGhgavfQ0NDQoLC2v37IkkORwOORyOQC8NAABYKuBnUNLS0lRRUeG1r7y8XGlpaYF+agAAcJXyO1CamppUU1OjmpoaSX/7GHFNTY3q6uok/e3HM/n5+Z7xDzzwgP785z9r8eLF+uijj/Tss8/q97//vX75y192zSsAAAC9jt+BUlVVpQkTJmjChAmSpMLCQk2YMEGPPvqoJKm+vt4TK5KUmJiorVu3qry8XCkpKVqxYoVefPFFPmIMAAB88vsalMzMTBljfN7f3m+JzczMVHV1tb9PBQAA+igrP8UDAAD6NgIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdToVKKtXr1ZCQoJCQ0M1ZcoUvfvuuz7HlpaWKigoyGsLDQ3t9IIBAEDv53egvPrqqyosLNTSpUt18OBBpaSkKDs7W6dPn/Y5JywsTPX19Z6ttrb2ihYNAAB6N78D5emnn9b8+fM1Z84cjR07VmvWrNGgQYO0du1an3OCgoIUExPj2aKjo69o0QAAoHfzK1DOnTunAwcOKCsr6+8PEBysrKwsVVZW+pzX1NSk+Ph4xcXFaebMmTpy5EiHz9PS0iKXy+W1AQCAvsOvQPn888/V2tra5gxIdHS0nE5nu3OSk5O1du1abd68Wb/97W/ldruVnp6uTz75xOfzlJSUKDw83LPFxcX5s0wAAHCVC/ineNLS0pSfn6/U1FRlZGTojTfe0NChQ/X888/7nFNUVKTGxkbPdurUqUAvEwAAWKS/P4OHDBmifv36qaGhwWt/Q0ODYmJiLusxBgwYoAkTJujYsWM+xzgcDjkcDn+WBgAAehG/zqCEhIRo4sSJqqio8Oxzu92qqKhQWlraZT1Ga2urDh06pNjYWP9WCgAA+gy/zqBIUmFhoWbPnq1JkyZp8uTJWrVqlZqbmzVnzhxJUn5+voYPH66SkhJJ0mOPPaapU6cqKSlJZ86c0fLly1VbW6t58+Z17SsBAAC9ht+BMmvWLH322Wd69NFH5XQ6lZqaqu3bt3sunK2rq1Nw8N9PzHz55ZeaP3++nE6nrr32Wk2cOFF79+7V2LFju+5VAACAXsXvQJGkgoICFRQUtHvfzp07vW6vXLlSK1eu7MzTAACAPoq/xQMAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6nQqU1atXKyEhQaGhoZoyZYrefffdDsdv2LBBo0ePVmhoqMaPH69t27Z1arEAAKBv8DtQXn31VRUWFmrp0qU6ePCgUlJSlJ2drdOnT7c7fu/evcrLy9PcuXNVXV2t3Nxc5ebm6vDhw1e8eAAA0Dv5HShPP/205s+frzlz5mjs2LFas2aNBg0apLVr17Y7/plnntEdd9yhhx9+WGPGjNGyZct0880367/+67+uePEAAKB36u/P4HPnzunAgQMqKiry7AsODlZWVpYqKyvbnVNZWanCwkKvfdnZ2dq0aZPP52lpaVFLS4vndmNjoyTJ5XL5s9zL4m75qssfM9ACcRwC7Wo8zkBHrsb/Dq9GfO3oHoF6P198XGOM33P9CpTPP/9cra2tio6O9tofHR2tjz76qN05Tqez3fFOp9Pn85SUlKi4uLjN/ri4OH+W22uFr+rpFQDgv0P0JoF+P589e1bh4eF+zfErULpLUVGR11kXt9utL774Qv/wD/+goKCgHlxZ4LlcLsXFxenUqVMKCwvr6eVYh+PTMY6PbxybjnF8Osbx6Ziv42OM0dmzZzVs2DC/H9OvQBkyZIj69eunhoYGr/0NDQ2KiYlpd05MTIxf4yXJ4XDI4XB47YuIiPBnqVe9sLAw/iPoAMenYxwf3zg2HeP4dIzj07H2jo+/Z04u8usi2ZCQEE2cOFEVFRWefW63WxUVFUpLS2t3Tlpamtd4SSovL/c5HgAAwO8f8RQWFmr27NmaNGmSJk+erFWrVqm5uVlz5syRJOXn52v48OEqKSmRJC1cuFAZGRlasWKFcnJytH79elVVVemFF17o2lcCAAB6Db8DZdasWfrss8/06KOPyul0KjU1Vdu3b/dcCFtXV6fg4L+fmElPT9e6dev0q1/9So888oiuv/56bdq0SePGjeu6V9GLOBwOLV26tM2PuPA3HJ+OcXx849h0jOPTMY5PxwJxfIJMZz77AwAAEED8LR4AAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CxQK//vWvlZ6erkGDBl32b8w1xujRRx9VbGysBg4cqKysLH388ceBXWgP+eKLL3TvvfcqLCxMERERmjt3rpqamjqck5mZqaCgIK/tgQce6KYVB9bq1auVkJCg0NBQTZkyRe+++26H4zds2KDRo0crNDRU48eP17Zt27pppd3Pn2NTWlra5j0SGhrajavtXrt379add96pYcOGKSgoqMM/2HrRzp07dfPNN8vhcCgpKUmlpaUBX2dP8PfY7Ny5s817JygoqMO/MXc1Kykp0S233KLBgwcrKipKubm5Onr06CXnXenXHgLFAufOndPdd9+tBx988LLnPPnkk/qP//gPrVmzRvv379d3vvMdZWdn65tvvgngSnvGvffeqyNHjqi8vFxbtmzR7t279bOf/eyS8+bPn6/6+nrP9uSTT3bDagPr1VdfVWFhoZYuXaqDBw8qJSVF2dnZOn36dLvj9+7dq7y8PM2dO1fV1dXKzc1Vbm6uDh8+3M0rDzx/j430t1/L/X/fI7W1td244u7V3NyslJQUrV69+rLGnzhxQjk5OZo+fbpqamq0aNEizZs3T2VlZQFeaffz99hcdPToUa/3T1RUVIBW2LN27dqlBQsWaN++fSovL9f58+d1++23q7m52eecLvnaY2CNl156yYSHh19ynNvtNjExMWb58uWefWfOnDEOh8P87ne/C+AKu98HH3xgJJn33nvPs+/NN980QUFB5tNPP/U5LyMjwyxcuLAbVti9Jk+ebBYsWOC53draaoYNG2ZKSkraHX/PPfeYnJwcr31Tpkwx999/f0DX2RP8PTaX+99bbyTJbNy4scMxixcvNjfeeKPXvlmzZpns7OwArqznXc6xefvtt40k8+WXX3bLmmxz+vRpI8ns2rXL55iu+NrDGZSr0IkTJ+R0OpWVleXZFx4erilTpqiysrIHV9b1KisrFRERoUmTJnn2ZWVlKTg4WPv37+9w7v/8z/9oyJAhGjdunIqKivTVV18FerkBde7cOR04cMDr3z04OFhZWVk+/90rKyu9xktSdnZ2r3ufdObYSFJTU5Pi4+MVFxenmTNn6siRI92x3KtCX3nvXInU1FTFxsbqtttu0zvvvNPTy+k2jY2NkqTIyEifY7ri/eP3r7pHz7v4c86Lf17goujo6F73M1Cn09nmtGn//v0VGRnZ4Wv98Y9/rPj4eA0bNkzvv/++/u3f/k1Hjx7VG2+8EeglB8znn3+u1tbWdv/dP/roo3bnOJ3OPvE+6cyxSU5O1tq1a3XTTTepsbFRTz31lNLT03XkyBGNGDGiO5ZtNV/vHZfLpa+//loDBw7soZX1vNjYWK1Zs0aTJk1SS0uLXnzxRWVmZmr//v26+eabe3p5AeV2u7Vo0SLdeuutHf7Jmq742kOgBMiSJUv0xBNPdDjmww8/1OjRo7tpRXa53OPTWf/3GpXx48crNjZW3//+93X8+HGNGjWq04+L3iMtLc3rr6qnp6drzJgxev7557Vs2bIeXBlsl5ycrOTkZM/t9PR0HT9+XCtXrtQrr7zSgysLvAULFujw4cPas2dPwJ+LQAmQhx56SPfdd1+HY6677rpOPXZMTIwkqaGhQbGxsZ79DQ0NSk1N7dRjdrfLPT4xMTFtLnK8cOGCvvjiC89xuBxTpkyRJB07duyqDZQhQ4aoX79+amho8Nrf0NDg81jExMT4Nf5q1Zlj820DBgzQhAkTdOzYsUAs8arj670TFhbWp8+e+DJ58uRu+abdkwoKCjwfVLjUWcau+NrDNSgBMnToUI0ePbrDLSQkpFOPnZiYqJiYGFVUVHj2uVwu7d+/3+v/CG12uccnLS1NZ86c0YEDBzxzd+zYIbfb7YmOy1FTUyNJXkF3tQkJCdHEiRO9/t3dbrcqKip8/runpaV5jZek8vLyq+Z9crk6c2y+rbW1VYcOHbqq3yNdqa+8d7pKTU1Nr33vGGNUUFCgjRs3aseOHUpMTLzknC55/3T2Kl50ndraWlNdXW2Ki4vNNddcY6qrq011dbU5e/asZ0xycrJ54403PLcff/xxExERYTZv3mzef/99M3PmTJOYmGi+/vrrnngJAXXHHXeYCRMmmP3795s9e/aY66+/3uTl5Xnu/+STT0xycrLZv3+/McaYY8eOmccee8xUVVWZEydOmM2bN5vrrrvOTJs2radeQpdZv369cTgcprS01HzwwQfmZz/7mYmIiDBOp9MYY8xPfvITs2TJEs/4d955x/Tv39889dRT5sMPPzRLly41AwYMMIcOHeqplxAw/h6b4uJiU1ZWZo4fP24OHDhgfvSjH5nQ0FBz5MiRnnoJAXX27FnP1xZJ5umnnzbV1dWmtrbWGGPMkiVLzE9+8hPP+D//+c9m0KBB5uGHHzYffvihWb16tenXr5/Zvn17T72EgPH32KxcudJs2rTJfPzxx+bQoUNm4cKFJjg42Lz11ls99RIC6sEHHzTh4eFm586dpr6+3rN99dVXnjGB+NpDoFhg9uzZRlKb7e233/aMkWReeuklz223223+/d//3URHRxuHw2G+//3vm6NHj3b/4rvBX//6V5OXl2euueYaExYWZubMmeMVbydOnPA6XnV1dWbatGkmMjLSOBwOk5SUZB5++GHT2NjYQ6+ga/3nf/6nGTlypAkJCTGTJ082+/bt89yXkZFhZs+e7TX+97//vbnhhhtMSEiIufHGG83WrVu7ecXdx59js2jRIs/Y6Oho84Mf/MAcPHiwB1bdPS5+NPbb28VjMnv2bJORkdFmTmpqqgkJCTHXXXed19eg3sTfY/PEE0+YUaNGmdDQUBMZGWkyMzPNjh07embx3aC9Y/Pt70mB+NoT9P+fHAAAwBpcgwIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6/w9O7SBFcq80UQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "As we can see on the x-axis in the above images, z-score results in our data having a mean of zero while min-max results in our data spanning values from 0 to +1, with both methods not changing the actual distribution of our data."
      ],
      "metadata": {
        "id": "w050t2twGGys"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## **Supervised Learning Regression Example**"
      ],
      "metadata": {
        "id": "JBirR3cdICH8"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Now, we will proceed with training a neural network in a supervised learning regression application."
      ],
      "metadata": {
        "id": "rT22AWjwIIeX"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "We will use a dataset from Molina et al. (2021) consisting of thunderstorms generated using a 4-km horizontal resolution WRF simulation. We are using a small portion of the dataset, specifically thunderstorms simulated for April 27, 2011 over the continental United States."
      ],
      "metadata": {
        "id": "Z6MIe4xpeOQq"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!wget https://zenodo.org/records/10576726/files/current_SP3hourly_20110427.nc"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YIDtIzo3IJFS",
        "outputId": "ae1c41ec-2e65-4a77-d390-a6442e7e6ce4"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2024-01-28 08:21:42--  https://zenodo.org/records/10576726/files/current_SP3hourly_20110427.nc\n",
            "Resolving zenodo.org (zenodo.org)... 188.184.98.238, 188.185.79.172, 188.184.103.159, ...\n",
            "Connecting to zenodo.org (zenodo.org)|188.184.98.238|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 109400556 (104M) [application/octet-stream]\n",
            "Saving to: ‘current_SP3hourly_20110427.nc.2’\n",
            "\n",
            "current_SP3hourly_2 100%[===================>] 104.33M  17.7MB/s    in 7.4s    \n",
            "\n",
            "2024-01-28 08:21:51 (14.1 MB/s) - ‘current_SP3hourly_20110427.nc.2’ saved [109400556/109400556]\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's visualize the dataset. We can use python libraries for this."
      ],
      "metadata": {
        "id": "FTRe0rEpexSJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import xarray as xr"
      ],
      "metadata": {
        "id": "V-_wY-Lqewnh"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "ds = xr.open_dataset('current_SP3hourly_20110427.nc')"
      ],
      "metadata": {
        "id": "pY1Bn-N8d7pB"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "ds"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 771
        },
        "id": "YYoMRMW9ewrH",
        "outputId": "e397f3f1-90de-4a9e-cb04-317c1f030dcd"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<xarray.Dataset>\n",
              "Dimensions:      (starttime: 2224, y: 32, x: 32)\n",
              "Coordinates:\n",
              "  * starttime    (starttime) float64 9.264e+04 9.264e+04 ... 9.266e+04 9.266e+04\n",
              "    endtime      (starttime) float64 ...\n",
              "    x_speed      (starttime) int64 ...\n",
              "    y_speed      (starttime) int64 ...\n",
              "Dimensions without coordinates: y, x\n",
              "Data variables:\n",
              "    uh25_grid    (starttime, y, x) float32 ...\n",
              "    uh03_grid    (starttime, y, x) float32 ...\n",
              "    ctt_grid     (starttime, y, x) float32 ...\n",
              "    dbz_grid     (starttime, y, x) float32 ...\n",
              "    mask         (starttime, y, x) int64 ...\n",
              "    row_indices  (starttime, y, x) int64 ...\n",
              "    col_indices  (starttime, y, x) int64 ...\n",
              "    lats         (starttime, y, x) float32 ...\n",
              "    lons         (starttime, y, x) float32 ..."
            ],
            "text/html": [
              "<div><svg style=\"position: absolute; width: 0; height: 0; overflow: hidden\">\n",
              "<defs>\n",
              "<symbol id=\"icon-database\" viewBox=\"0 0 32 32\">\n",
              "<path d=\"M16 0c-8.837 0-16 2.239-16 5v4c0 2.761 7.163 5 16 5s16-2.239 16-5v-4c0-2.761-7.163-5-16-5z\"></path>\n",
              "<path d=\"M16 17c-8.837 0-16-2.239-16-5v6c0 2.761 7.163 5 16 5s16-2.239 16-5v-6c0 2.761-7.163 5-16 5z\"></path>\n",
              "<path d=\"M16 26c-8.837 0-16-2.239-16-5v6c0 2.761 7.163 5 16 5s16-2.239 16-5v-6c0 2.761-7.163 5-16 5z\"></path>\n",
              "</symbol>\n",
              "<symbol id=\"icon-file-text2\" viewBox=\"0 0 32 32\">\n",
              "<path d=\"M28.681 7.159c-0.694-0.947-1.662-2.053-2.724-3.116s-2.169-2.030-3.116-2.724c-1.612-1.182-2.393-1.319-2.841-1.319h-15.5c-1.378 0-2.5 1.121-2.5 2.5v27c0 1.378 1.122 2.5 2.5 2.5h23c1.378 0 2.5-1.122 2.5-2.5v-19.5c0-0.448-0.137-1.23-1.319-2.841zM24.543 5.457c0.959 0.959 1.712 1.825 2.268 2.543h-4.811v-4.811c0.718 0.556 1.584 1.309 2.543 2.268zM28 29.5c0 0.271-0.229 0.5-0.5 0.5h-23c-0.271 0-0.5-0.229-0.5-0.5v-27c0-0.271 0.229-0.5 0.5-0.5 0 0 15.499-0 15.5 0v7c0 0.552 0.448 1 1 1h7v19.5z\"></path>\n",
              "<path d=\"M23 26h-14c-0.552 0-1-0.448-1-1s0.448-1 1-1h14c0.552 0 1 0.448 1 1s-0.448 1-1 1z\"></path>\n",
              "<path d=\"M23 22h-14c-0.552 0-1-0.448-1-1s0.448-1 1-1h14c0.552 0 1 0.448 1 1s-0.448 1-1 1z\"></path>\n",
              "<path d=\"M23 18h-14c-0.552 0-1-0.448-1-1s0.448-1 1-1h14c0.552 0 1 0.448 1 1s-0.448 1-1 1z\"></path>\n",
              "</symbol>\n",
              "</defs>\n",
              "</svg>\n",
              "<style>/* CSS stylesheet for displaying xarray objects in jupyterlab.\n",
              " *\n",
              " */\n",
              "\n",
              ":root {\n",
              "  --xr-font-color0: var(--jp-content-font-color0, rgba(0, 0, 0, 1));\n",
              "  --xr-font-color2: var(--jp-content-font-color2, rgba(0, 0, 0, 0.54));\n",
              "  --xr-font-color3: var(--jp-content-font-color3, rgba(0, 0, 0, 0.38));\n",
              "  --xr-border-color: var(--jp-border-color2, #e0e0e0);\n",
              "  --xr-disabled-color: var(--jp-layout-color3, #bdbdbd);\n",
              "  --xr-background-color: var(--jp-layout-color0, white);\n",
              "  --xr-background-color-row-even: var(--jp-layout-color1, white);\n",
              "  --xr-background-color-row-odd: var(--jp-layout-color2, #eeeeee);\n",
              "}\n",
              "\n",
              "html[theme=dark],\n",
              "body[data-theme=dark],\n",
              "body.vscode-dark {\n",
              "  --xr-font-color0: rgba(255, 255, 255, 1);\n",
              "  --xr-font-color2: rgba(255, 255, 255, 0.54);\n",
              "  --xr-font-color3: rgba(255, 255, 255, 0.38);\n",
              "  --xr-border-color: #1F1F1F;\n",
              "  --xr-disabled-color: #515151;\n",
              "  --xr-background-color: #111111;\n",
              "  --xr-background-color-row-even: #111111;\n",
              "  --xr-background-color-row-odd: #313131;\n",
              "}\n",
              "\n",
              ".xr-wrap {\n",
              "  display: block !important;\n",
              "  min-width: 300px;\n",
              "  max-width: 700px;\n",
              "}\n",
              "\n",
              ".xr-text-repr-fallback {\n",
              "  /* fallback to plain text repr when CSS is not injected (untrusted notebook) */\n",
              "  display: none;\n",
              "}\n",
              "\n",
              ".xr-header {\n",
              "  padding-top: 6px;\n",
              "  padding-bottom: 6px;\n",
              "  margin-bottom: 4px;\n",
              "  border-bottom: solid 1px var(--xr-border-color);\n",
              "}\n",
              "\n",
              ".xr-header > div,\n",
              ".xr-header > ul {\n",
              "  display: inline;\n",
              "  margin-top: 0;\n",
              "  margin-bottom: 0;\n",
              "}\n",
              "\n",
              ".xr-obj-type,\n",
              ".xr-array-name {\n",
              "  margin-left: 2px;\n",
              "  margin-right: 10px;\n",
              "}\n",
              "\n",
              ".xr-obj-type {\n",
              "  color: var(--xr-font-color2);\n",
              "}\n",
              "\n",
              ".xr-sections {\n",
              "  padding-left: 0 !important;\n",
              "  display: grid;\n",
              "  grid-template-columns: 150px auto auto 1fr 20px 20px;\n",
              "}\n",
              "\n",
              ".xr-section-item {\n",
              "  display: contents;\n",
              "}\n",
              "\n",
              ".xr-section-item input {\n",
              "  display: none;\n",
              "}\n",
              "\n",
              ".xr-section-item input + label {\n",
              "  color: var(--xr-disabled-color);\n",
              "}\n",
              "\n",
              ".xr-section-item input:enabled + label {\n",
              "  cursor: pointer;\n",
              "  color: var(--xr-font-color2);\n",
              "}\n",
              "\n",
              ".xr-section-item input:enabled + label:hover {\n",
              "  color: var(--xr-font-color0);\n",
              "}\n",
              "\n",
              ".xr-section-summary {\n",
              "  grid-column: 1;\n",
              "  color: var(--xr-font-color2);\n",
              "  font-weight: 500;\n",
              "}\n",
              "\n",
              ".xr-section-summary > span {\n",
              "  display: inline-block;\n",
              "  padding-left: 0.5em;\n",
              "}\n",
              "\n",
              ".xr-section-summary-in:disabled + label {\n",
              "  color: var(--xr-font-color2);\n",
              "}\n",
              "\n",
              ".xr-section-summary-in + label:before {\n",
              "  display: inline-block;\n",
              "  content: '►';\n",
              "  font-size: 11px;\n",
              "  width: 15px;\n",
              "  text-align: center;\n",
              "}\n",
              "\n",
              ".xr-section-summary-in:disabled + label:before {\n",
              "  color: var(--xr-disabled-color);\n",
              "}\n",
              "\n",
              ".xr-section-summary-in:checked + label:before {\n",
              "  content: '▼';\n",
              "}\n",
              "\n",
              ".xr-section-summary-in:checked + label > span {\n",
              "  display: none;\n",
              "}\n",
              "\n",
              ".xr-section-summary,\n",
              ".xr-section-inline-details {\n",
              "  padding-top: 4px;\n",
              "  padding-bottom: 4px;\n",
              "}\n",
              "\n",
              ".xr-section-inline-details {\n",
              "  grid-column: 2 / -1;\n",
              "}\n",
              "\n",
              ".xr-section-details {\n",
              "  display: none;\n",
              "  grid-column: 1 / -1;\n",
              "  margin-bottom: 5px;\n",
              "}\n",
              "\n",
              ".xr-section-summary-in:checked ~ .xr-section-details {\n",
              "  display: contents;\n",
              "}\n",
              "\n",
              ".xr-array-wrap {\n",
              "  grid-column: 1 / -1;\n",
              "  display: grid;\n",
              "  grid-template-columns: 20px auto;\n",
              "}\n",
              "\n",
              ".xr-array-wrap > label {\n",
              "  grid-column: 1;\n",
              "  vertical-align: top;\n",
              "}\n",
              "\n",
              ".xr-preview {\n",
              "  color: var(--xr-font-color3);\n",
              "}\n",
              "\n",
              ".xr-array-preview,\n",
              ".xr-array-data {\n",
              "  padding: 0 5px !important;\n",
              "  grid-column: 2;\n",
              "}\n",
              "\n",
              ".xr-array-data,\n",
              ".xr-array-in:checked ~ .xr-array-preview {\n",
              "  display: none;\n",
              "}\n",
              "\n",
              ".xr-array-in:checked ~ .xr-array-data,\n",
              ".xr-array-preview {\n",
              "  display: inline-block;\n",
              "}\n",
              "\n",
              ".xr-dim-list {\n",
              "  display: inline-block !important;\n",
              "  list-style: none;\n",
              "  padding: 0 !important;\n",
              "  margin: 0;\n",
              "}\n",
              "\n",
              ".xr-dim-list li {\n",
              "  display: inline-block;\n",
              "  padding: 0;\n",
              "  margin: 0;\n",
              "}\n",
              "\n",
              ".xr-dim-list:before {\n",
              "  content: '(';\n",
              "}\n",
              "\n",
              ".xr-dim-list:after {\n",
              "  content: ')';\n",
              "}\n",
              "\n",
              ".xr-dim-list li:not(:last-child):after {\n",
              "  content: ',';\n",
              "  padding-right: 5px;\n",
              "}\n",
              "\n",
              ".xr-has-index {\n",
              "  font-weight: bold;\n",
              "}\n",
              "\n",
              ".xr-var-list,\n",
              ".xr-var-item {\n",
              "  display: contents;\n",
              "}\n",
              "\n",
              ".xr-var-item > div,\n",
              ".xr-var-item label,\n",
              ".xr-var-item > .xr-var-name span {\n",
              "  background-color: var(--xr-background-color-row-even);\n",
              "  margin-bottom: 0;\n",
              "}\n",
              "\n",
              ".xr-var-item > .xr-var-name:hover span {\n",
              "  padding-right: 5px;\n",
              "}\n",
              "\n",
              ".xr-var-list > li:nth-child(odd) > div,\n",
              ".xr-var-list > li:nth-child(odd) > label,\n",
              ".xr-var-list > li:nth-child(odd) > .xr-var-name span {\n",
              "  background-color: var(--xr-background-color-row-odd);\n",
              "}\n",
              "\n",
              ".xr-var-name {\n",
              "  grid-column: 1;\n",
              "}\n",
              "\n",
              ".xr-var-dims {\n",
              "  grid-column: 2;\n",
              "}\n",
              "\n",
              ".xr-var-dtype {\n",
              "  grid-column: 3;\n",
              "  text-align: right;\n",
              "  color: var(--xr-font-color2);\n",
              "}\n",
              "\n",
              ".xr-var-preview {\n",
              "  grid-column: 4;\n",
              "}\n",
              "\n",
              ".xr-index-preview {\n",
              "  grid-column: 2 / 5;\n",
              "  color: var(--xr-font-color2);\n",
              "}\n",
              "\n",
              ".xr-var-name,\n",
              ".xr-var-dims,\n",
              ".xr-var-dtype,\n",
              ".xr-preview,\n",
              ".xr-attrs dt {\n",
              "  white-space: nowrap;\n",
              "  overflow: hidden;\n",
              "  text-overflow: ellipsis;\n",
              "  padding-right: 10px;\n",
              "}\n",
              "\n",
              ".xr-var-name:hover,\n",
              ".xr-var-dims:hover,\n",
              ".xr-var-dtype:hover,\n",
              ".xr-attrs dt:hover {\n",
              "  overflow: visible;\n",
              "  width: auto;\n",
              "  z-index: 1;\n",
              "}\n",
              "\n",
              ".xr-var-attrs,\n",
              ".xr-var-data,\n",
              ".xr-index-data {\n",
              "  display: none;\n",
              "  background-color: var(--xr-background-color) !important;\n",
              "  padding-bottom: 5px !important;\n",
              "}\n",
              "\n",
              ".xr-var-attrs-in:checked ~ .xr-var-attrs,\n",
              ".xr-var-data-in:checked ~ .xr-var-data,\n",
              ".xr-index-data-in:checked ~ .xr-index-data {\n",
              "  display: block;\n",
              "}\n",
              "\n",
              ".xr-var-data > table {\n",
              "  float: right;\n",
              "}\n",
              "\n",
              ".xr-var-name span,\n",
              ".xr-var-data,\n",
              ".xr-index-name div,\n",
              ".xr-index-data,\n",
              ".xr-attrs {\n",
              "  padding-left: 25px !important;\n",
              "}\n",
              "\n",
              ".xr-attrs,\n",
              ".xr-var-attrs,\n",
              ".xr-var-data,\n",
              ".xr-index-data {\n",
              "  grid-column: 1 / -1;\n",
              "}\n",
              "\n",
              "dl.xr-attrs {\n",
              "  padding: 0;\n",
              "  margin: 0;\n",
              "  display: grid;\n",
              "  grid-template-columns: 125px auto;\n",
              "}\n",
              "\n",
              ".xr-attrs dt,\n",
              ".xr-attrs dd {\n",
              "  padding: 0;\n",
              "  margin: 0;\n",
              "  float: left;\n",
              "  padding-right: 10px;\n",
              "  width: auto;\n",
              "}\n",
              "\n",
              ".xr-attrs dt {\n",
              "  font-weight: normal;\n",
              "  grid-column: 1;\n",
              "}\n",
              "\n",
              ".xr-attrs dt:hover span {\n",
              "  display: inline-block;\n",
              "  background: var(--xr-background-color);\n",
              "  padding-right: 10px;\n",
              "}\n",
              "\n",
              ".xr-attrs dd {\n",
              "  grid-column: 2;\n",
              "  white-space: pre-wrap;\n",
              "  word-break: break-all;\n",
              "}\n",
              "\n",
              ".xr-icon-database,\n",
              ".xr-icon-file-text2,\n",
              ".xr-no-icon {\n",
              "  display: inline-block;\n",
              "  vertical-align: middle;\n",
              "  width: 1em;\n",
              "  height: 1.5em !important;\n",
              "  stroke-width: 0;\n",
              "  stroke: currentColor;\n",
              "  fill: currentColor;\n",
              "}\n",
              "</style><pre class='xr-text-repr-fallback'>&lt;xarray.Dataset&gt;\n",
              "Dimensions:      (starttime: 2224, y: 32, x: 32)\n",
              "Coordinates:\n",
              "  * starttime    (starttime) float64 9.264e+04 9.264e+04 ... 9.266e+04 9.266e+04\n",
              "    endtime      (starttime) float64 ...\n",
              "    x_speed      (starttime) int64 ...\n",
              "    y_speed      (starttime) int64 ...\n",
              "Dimensions without coordinates: y, x\n",
              "Data variables:\n",
              "    uh25_grid    (starttime, y, x) float32 ...\n",
              "    uh03_grid    (starttime, y, x) float32 ...\n",
              "    ctt_grid     (starttime, y, x) float32 ...\n",
              "    dbz_grid     (starttime, y, x) float32 ...\n",
              "    mask         (starttime, y, x) int64 ...\n",
              "    row_indices  (starttime, y, x) int64 ...\n",
              "    col_indices  (starttime, y, x) int64 ...\n",
              "    lats         (starttime, y, x) float32 ...\n",
              "    lons         (starttime, y, x) float32 ...</pre><div class='xr-wrap' style='display:none'><div class='xr-header'><div class='xr-obj-type'>xarray.Dataset</div></div><ul class='xr-sections'><li class='xr-section-item'><input id='section-b082ab13-3134-448e-b844-31b033f7893c' class='xr-section-summary-in' type='checkbox' disabled ><label for='section-b082ab13-3134-448e-b844-31b033f7893c' class='xr-section-summary'  title='Expand/collapse section'>Dimensions:</label><div class='xr-section-inline-details'><ul class='xr-dim-list'><li><span class='xr-has-index'>starttime</span>: 2224</li><li><span>y</span>: 32</li><li><span>x</span>: 32</li></ul></div><div class='xr-section-details'></div></li><li class='xr-section-item'><input id='section-05db0b48-8a13-4ab0-8408-dca9bcd9403f' class='xr-section-summary-in' type='checkbox'  checked><label for='section-05db0b48-8a13-4ab0-8408-dca9bcd9403f' class='xr-section-summary' >Coordinates: <span>(4)</span></label><div class='xr-section-inline-details'></div><div class='xr-section-details'><ul class='xr-var-list'><li class='xr-var-item'><div class='xr-var-name'><span class='xr-has-index'>starttime</span></div><div class='xr-var-dims'>(starttime)</div><div class='xr-var-dtype'>float64</div><div class='xr-var-preview xr-preview'>9.264e+04 9.264e+04 ... 9.266e+04</div><input id='attrs-1cea445a-8258-4701-aab5-98e77f942663' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-1cea445a-8258-4701-aab5-98e77f942663' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-43ecc2db-8ee9-42b2-8f0d-bb128429a410' class='xr-var-data-in' type='checkbox'><label for='data-43ecc2db-8ee9-42b2-8f0d-bb128429a410' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>array([92640., 92640., 92640., ..., 92661., 92661., 92661.])</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>endtime</span></div><div class='xr-var-dims'>(starttime)</div><div class='xr-var-dtype'>float64</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-d1a9a773-1403-4b2b-be74-a9cb5ffe146d' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-d1a9a773-1403-4b2b-be74-a9cb5ffe146d' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-47049e69-24d2-491d-8907-943ae69895f7' class='xr-var-data-in' type='checkbox'><label for='data-47049e69-24d2-491d-8907-943ae69895f7' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2224 values with dtype=float64]</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>x_speed</span></div><div class='xr-var-dims'>(starttime)</div><div class='xr-var-dtype'>int64</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-486859a8-8cc7-42f6-a81c-a218d4074ac9' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-486859a8-8cc7-42f6-a81c-a218d4074ac9' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-520f3697-6cb6-40c3-b170-55544dea088a' class='xr-var-data-in' type='checkbox'><label for='data-520f3697-6cb6-40c3-b170-55544dea088a' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2224 values with dtype=int64]</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>y_speed</span></div><div class='xr-var-dims'>(starttime)</div><div class='xr-var-dtype'>int64</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-82649973-9f80-4416-a162-18e6b639a723' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-82649973-9f80-4416-a162-18e6b639a723' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-fb899fcc-d2cf-4d2c-8599-0c43e4726e82' class='xr-var-data-in' type='checkbox'><label for='data-fb899fcc-d2cf-4d2c-8599-0c43e4726e82' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2224 values with dtype=int64]</pre></div></li></ul></div></li><li class='xr-section-item'><input id='section-bae2bf39-6dfa-4778-905a-2c4df84f656a' class='xr-section-summary-in' type='checkbox'  checked><label for='section-bae2bf39-6dfa-4778-905a-2c4df84f656a' class='xr-section-summary' >Data variables: <span>(9)</span></label><div class='xr-section-inline-details'></div><div class='xr-section-details'><ul class='xr-var-list'><li class='xr-var-item'><div class='xr-var-name'><span>uh25_grid</span></div><div class='xr-var-dims'>(starttime, y, x)</div><div class='xr-var-dtype'>float32</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-5d41c05b-1c1b-4a59-8436-ce7b703b9c39' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-5d41c05b-1c1b-4a59-8436-ce7b703b9c39' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-06f75f82-4051-40af-8053-94d69e277c8f' class='xr-var-data-in' type='checkbox'><label for='data-06f75f82-4051-40af-8053-94d69e277c8f' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2277376 values with dtype=float32]</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>uh03_grid</span></div><div class='xr-var-dims'>(starttime, y, x)</div><div class='xr-var-dtype'>float32</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-4d2dd136-99db-4d44-9367-ba87615cffa7' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-4d2dd136-99db-4d44-9367-ba87615cffa7' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-2da93822-a2ef-4114-aa25-b4b4e8e2bff3' class='xr-var-data-in' type='checkbox'><label for='data-2da93822-a2ef-4114-aa25-b4b4e8e2bff3' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2277376 values with dtype=float32]</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>ctt_grid</span></div><div class='xr-var-dims'>(starttime, y, x)</div><div class='xr-var-dtype'>float32</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-c5a76292-8108-4c17-aea1-214f6bcdf38a' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-c5a76292-8108-4c17-aea1-214f6bcdf38a' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-2c5fc053-51fc-44dd-aa2b-7d14874e99a7' class='xr-var-data-in' type='checkbox'><label for='data-2c5fc053-51fc-44dd-aa2b-7d14874e99a7' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2277376 values with dtype=float32]</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>dbz_grid</span></div><div class='xr-var-dims'>(starttime, y, x)</div><div class='xr-var-dtype'>float32</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-606324e0-bbf7-4951-9620-f953386962d0' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-606324e0-bbf7-4951-9620-f953386962d0' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-84a4d13a-5454-41aa-ab87-836fdd9a68df' class='xr-var-data-in' type='checkbox'><label for='data-84a4d13a-5454-41aa-ab87-836fdd9a68df' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2277376 values with dtype=float32]</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>mask</span></div><div class='xr-var-dims'>(starttime, y, x)</div><div class='xr-var-dtype'>int64</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-c36ebcf3-70d7-4c95-b996-ee535d4fda34' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-c36ebcf3-70d7-4c95-b996-ee535d4fda34' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-4edee7b3-1b1d-4bb7-b568-b1de7f91a195' class='xr-var-data-in' type='checkbox'><label for='data-4edee7b3-1b1d-4bb7-b568-b1de7f91a195' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2277376 values with dtype=int64]</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>row_indices</span></div><div class='xr-var-dims'>(starttime, y, x)</div><div class='xr-var-dtype'>int64</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-f10d7b3c-b80e-4aed-a6a2-802706263963' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-f10d7b3c-b80e-4aed-a6a2-802706263963' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-cb9658cb-0454-404d-84c2-7483d0756163' class='xr-var-data-in' type='checkbox'><label for='data-cb9658cb-0454-404d-84c2-7483d0756163' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2277376 values with dtype=int64]</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>col_indices</span></div><div class='xr-var-dims'>(starttime, y, x)</div><div class='xr-var-dtype'>int64</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-b086f593-2b1e-435e-a8a0-11c663312f9a' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-b086f593-2b1e-435e-a8a0-11c663312f9a' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-64b0b6a5-c66f-4521-b798-9dceafebe3d9' class='xr-var-data-in' type='checkbox'><label for='data-64b0b6a5-c66f-4521-b798-9dceafebe3d9' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2277376 values with dtype=int64]</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>lats</span></div><div class='xr-var-dims'>(starttime, y, x)</div><div class='xr-var-dtype'>float32</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-a642ad1a-4ad7-4e9c-a92f-5d5c767e4d4d' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-a642ad1a-4ad7-4e9c-a92f-5d5c767e4d4d' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-d9bb6209-a175-4859-803c-58ff18a386cc' class='xr-var-data-in' type='checkbox'><label for='data-d9bb6209-a175-4859-803c-58ff18a386cc' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2277376 values with dtype=float32]</pre></div></li><li class='xr-var-item'><div class='xr-var-name'><span>lons</span></div><div class='xr-var-dims'>(starttime, y, x)</div><div class='xr-var-dtype'>float32</div><div class='xr-var-preview xr-preview'>...</div><input id='attrs-9f35d7c4-66b1-43cb-af00-33d4f94f0b18' class='xr-var-attrs-in' type='checkbox' disabled><label for='attrs-9f35d7c4-66b1-43cb-af00-33d4f94f0b18' title='Show/Hide attributes'><svg class='icon xr-icon-file-text2'><use xlink:href='#icon-file-text2'></use></svg></label><input id='data-32b69820-f523-4951-8802-c7ce26f58052' class='xr-var-data-in' type='checkbox'><label for='data-32b69820-f523-4951-8802-c7ce26f58052' title='Show/Hide data repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-var-attrs'><dl class='xr-attrs'></dl></div><div class='xr-var-data'><pre>[2277376 values with dtype=float32]</pre></div></li></ul></div></li><li class='xr-section-item'><input id='section-addcb367-5c64-42d4-abb3-f5e16df40e0f' class='xr-section-summary-in' type='checkbox'  ><label for='section-addcb367-5c64-42d4-abb3-f5e16df40e0f' class='xr-section-summary' >Indexes: <span>(1)</span></label><div class='xr-section-inline-details'></div><div class='xr-section-details'><ul class='xr-var-list'><li class='xr-var-item'><div class='xr-index-name'><div>starttime</div></div><div class='xr-index-preview'>PandasIndex</div><div></div><input id='index-d19d1242-eb15-44bb-a952-ae4dcfd9e106' class='xr-index-data-in' type='checkbox'/><label for='index-d19d1242-eb15-44bb-a952-ae4dcfd9e106' title='Show/Hide index repr'><svg class='icon xr-icon-database'><use xlink:href='#icon-database'></use></svg></label><div class='xr-index-data'><pre>PandasIndex(Float64Index([92640.0, 92640.0, 92640.0, 92640.0, 92640.0, 92640.0, 92640.0,\n",
              "              92640.0, 92640.0, 92640.0,\n",
              "              ...\n",
              "              92661.0, 92661.0, 92661.0, 92661.0, 92661.0, 92661.0, 92661.0,\n",
              "              92661.0, 92661.0, 92661.0],\n",
              "             dtype=&#x27;float64&#x27;, name=&#x27;starttime&#x27;, length=2224))</pre></div></li></ul></div></li><li class='xr-section-item'><input id='section-dd5a2541-2a50-43b5-9666-5250aa36972a' class='xr-section-summary-in' type='checkbox' disabled ><label for='section-dd5a2541-2a50-43b5-9666-5250aa36972a' class='xr-section-summary'  title='Expand/collapse section'>Attributes: <span>(0)</span></label><div class='xr-section-inline-details'></div><div class='xr-section-details'><dl class='xr-attrs'></dl></div></li></ul></div></div>"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "The dataset contains 2,224 storm samples that were generated using the github reposity [deep-conus](https://github.com/mariajmolina/deep-conus). Each sample is 32x32 grid cells (i.e., pixels) with each grid cell spanning 4-km.\n",
        "\n",
        "Let's visualize some of the variables."
      ],
      "metadata": {
        "id": "yw6ewfZee-lr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# simulated radar reflectivity\n",
        "ds['dbz_grid'].isel(starttime=0).plot()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "DsoCCgnSfnhJ",
        "outputId": "c5b69b9a-dbf9-418c-f71e-3bff8ee22e86"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.QuadMesh at 0x7f0a98f51690>"
            ]
          },
          "metadata": {},
          "execution_count": 9
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# gridded updraft helicity from 2-5 km\n",
        "ds['uh25_grid'].isel(starttime=0).plot()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "bFIWAsD0ewu-",
        "outputId": "52829061-c6dc-40af-dc93-ccc2ad462f36"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.QuadMesh at 0x7f0a98e49c00>"
            ]
          },
          "metadata": {},
          "execution_count": 10
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjUAAAHHCAYAAABHp6kXAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABS4klEQVR4nO3dd3hUZfo38O9MyEz6hJAOAUM3QkBQMDQpMbR1Q1EQWQmCoGxAAUWJ0hXisqwgbCgiBtYVCyigqKwY2roEWCL8KAoCBhNKAghkQiCFzP3+wZtZhkmbnEnOZPL9XNe5YJ455T5nSu55ynk0IiIgIiIiquW0agdAREREZA9MaoiIiMgpMKkhIiIip8CkhoiIiJwCkxoiIiJyCkxqiIiIyCkwqSEiIiKnwKSGiIiInAKTGiIiInIKTGoUOnv2LDQaDdauXat2KETV5r777sPo0aMrtW7Pnj3Rs2fPao2HiKg0tS6pWb9+PZYsWVIt+75w4QLmzJmDw4cP1+hxHd0nn3yCDh06wM3NDQEBARg7diyuXLlS4XY3b95EUlISYmJiEBISAm9vbzz44INYsWIFiouLS93mzJkzePrppxEYGAh3d3e0aNECb7zxRpnHKCoqQkREBDQaDRYtWlTlc1RizZo1uP/+++Hm5oYWLVpg2bJlFW7z2GOPQaPRYOLEiTUQof399NNPmDNnDs6ePat2KDXu9OnTeOKJJ1C/fn14eHigW7du2LlzZ6W2TUlJwZgxY9CyZUt4eHigadOmeO6553Dx4sVS1y8sLMSCBQvQunVruLm5ISgoCAMHDsS5c+fKPMb8+fOh0WjQpk2bKp2fUj///DP69esHLy8v+Pn54ZlnnsHly5fL3eajjz6CRqOBl5dXDUVJTktqmYEDB0qTJk2qZd///e9/BYAkJydX+rgmk0lu3bolt2/frpaY1LZ8+XIBIH369JGkpCRJSEgQDw8PiYyMlFu3bpW77dGjR0Wj0Uh0dLQsXLhQVq5cKYMHDxYAMmrUKKv1Dx06JAaDQSIiIuTtt9+W1atXy8yZM2X06NFlHuNvf/ubeHp6CgD561//qvh8bbVy5UoBIEOHDpX33ntPnnnmGQEgb7/9dpnbfP755+aY4+PjazDaqmvSpInExcWZH2/YsEEAyM6dO63WLSgokIKCgpoLrgZlZGSIv7+/BAUFyfz582XJkiXSrl07qVevnuzevbvC7Tt27Cjh4eHy6quvyurVqyUhIUG8vb0lKChILl68aLFuYWGhREdHi4eHh7z00kuyZs0aWbRokTz55JNy7NixUvefmZkpHh4e4unpKQ888IBdztkWmZmZ4u/vL82aNZN3331X5s+fL/Xr15d27dqV+Z7Izc2V0NBQ8fT0FE9PzxqOmJwNkxoRKSoqkoKCgiolNc6soKBAfH19pUePHmIymczlX331lQCQpUuXlrv95cuXS/3yffbZZwWAnDp1ylxWXFwsbdq0kc6dO8vNmzcrFV92drYYDAaZN2+e3ZOa5ORkqSjnv3nzpjRo0EAGDhxoUT5y5Ejx9PSUq1evWm1z69Ytue+++8wxO2NS48z+/Oc/S7169eTEiRPmsry8PAkLC5MOHTpUuP3u3buluLjYqgyAvPHGGxblf/nLX8TV1VX2799f6fiGDx8uvXv3lkcffdTuSU1Z3413mzBhgri7u8tvv/1mLtu+fbsAkFWrVpW6zWuvvSatWrUyf26IlHCopMZoNMpLL70kTZo0EZ1OJwEBARIdHS1paWkiIvLoo48KAIulJNEoKCiQmTNnSocOHcTHx0c8PDykW7dusmPHDotjpKenm/8ALl68WJo2bSparVYWL15ste+SD3F5xy3Z390f9ri4OPH09JTffvtNBg4cKJ6enhIaGip///vfRUTkyJEj0qtXL/Hw8JDGjRvLRx99ZHUtrl27Ji+99JI0atRIdDqdNGvWTN5++22rL8TqlJaWJgAkKSnJ6jkvLy/p0qVLlfb75ZdfCgD58ssvzWXffvutAJBvvvlGRO78oaio9uvZZ5+VTp06ya+//lpmUlPV61iZpObrr78WAPL1119blO/du1cAyIcffmi1zdy5c6Vx48Zy8+ZNuyQ1P//8swwdOlTq168ver1eOnbsKFu2bCn1XH744QeZMmWK+Pv7i4eHhwwaNEguXbpksa7JZJI333xTGjZsKO7u7tKzZ085duyYRVJTsr97l5IE59FHH5VHH33UvM+dO3cKAPn0009lzpw5EhoaKl5eXjJ06FC5fv265Ofny0svvSQBAQHi6ekpo0ePlvz8fKtz/fDDD6VDhw7i5uYm9evXl+HDh0tGRoai62ertm3bysMPP2xVHh8fLwDkl19+qdJ+/fz8ZMiQIebHxcXFEhoaKsOGDROROz+88vLyyt3H7t27xcXFRY4cOVJuUlPV61iZpCYwMFCefPJJq/KWLVtKnz59rMp/+eUX0el08vXXX5u/N6tq1KhRotfr5aeffrIoj4mJEV9fXzl//nyl9/Xdd99J165dxWAwiKenp7Rs2VISEhKqHBvVnHr2bs5S4oUXXsDGjRsxceJERERE4Pfff8cPP/yAn3/+GR06dMAbb7yBnJwcnDt3DosXLwYAcxus0WjE+++/jxEjRmDcuHHIzc3FmjVr0LdvXxw4cADt27e3OFZycjLy8/Mxfvx46PV6DB48GLm5uZg1axbGjx+P7t27AwC6dOmChg0blnncshQXF6N///7o0aMHFi5ciI8++ggTJ06Ep6cn3njjDYwcORJDhgzBypUrMWrUKERFRSE8PBzAnb4ojz76KM6fP4/nn38ejRs3xt69e5GQkICLFy9W2Lfnxo0byM/Pr/B6u7q6wmAwlPl8QUEBAMDd3d3qOXd3dxw6dAgmkwlarW1ds7KysgAA/v7+5rLvv/8eAKDX6/HQQw8hLS0NOp0OgwcPxvLly+Hn52exjwMHDmDdunX44YcfoNFoSj2O0utYkUOHDgEAHnroIYvyjh07QqvV4tChQ/jTn/5kLs/IyMDbb7+NDz74oNRraqvjx4+ja9euaNiwIaZPnw5PT0989tlnGDRoED7//HMMHjzYYv1Jkyahfv36mD17Ns6ePYslS5Zg4sSJ+PTTT83rzJo1C2+99RYGDBiAAQMG4Mcff0RMTAwKCwvN6/To0QMvvvgili5ditdffx33338/AJj/LUtiYiLc3d0xffp0nD59GsuWLYOrqyu0Wi2uXbuGOXPmYN++fVi7di3Cw8Mxa9Ys87bz58/HzJkzMWzYMDz33HO4fPkyli1bhh49euDQoUPw9fUt87hFRUXIycmp1DX18/Mr9/1cUFCA+vXrW5V7eHgAANLS0tCiRYtKHavEjRs3cOPGDYvPw08//YQLFy4gMjIS48ePx7p161BYWIi2bdvi3XffRa9evSz2UVxcjEmTJuG5555D27ZtyzyWkutYkfPnz+PSpUtWnwcA6NSpE7755hur8smTJ6NXr14YMGAAPvvssyofGwDeffdd7NixA3FxcUhNTYWLiwtWrVqF7777Dh9++CFCQ0MrtZ/jx4/jD3/4AyIjIzFv3jzo9XqcPn0a//nPfxTFRzVE7azqbgaDocJfrmU1A92+fduqzfbatWsSFBQkY8aMMZeV1Kz4+PhY/UqtSvNTWTU1AGTBggUWsbi7u4tGo5FPPvnEXH7ixAkBILNnzzaXvfnmm+Lp6Wn1q2/69Oni4uJS4a+qkuNXtNz9a7o0ly9fFo1GI2PHjrUoL4kZgFy5cqXcfdyroKBAIiIiJDw8XIqKiszlf/zjHwWANGjQQEaOHCkbN26UmTNnSr169aRLly4WzV8mk0k6deokI0aMEBHL2re7KbmOlampiY+PFxcXl1KfCwgIkKeeesqi7IknnrCo3YLCmpo+ffpI27ZtLWo1TCaTdOnSRVq0aGEuKzmX6Ohoi+s4ZcoUcXFxkevXr4uIyKVLl0Sn08nAgQMt1nv99dcFQKWbn8qqqWnTpo0UFhaay0eMGCEajUb69+9vsX1UVJTFZ+3s2bPi4uIi8+fPt1jv6NGjUq9ePavye5UcvzJLenp6uft6/PHHxdfXV4xGo1XMAGTRokXlbl+aN998UwBISkqKueyLL74wfx5atGghycnJkpycLC1atBCdTif/93//Z7GPv//972IwGMzfaaXV1Ci9jmV9N5Yo+f78xz/+YfXctGnTBIDFe3Xr1q1Sr149OX78uIiI4poaEZF//etfAkDeeust+fXXX8XLy0sGDRpk0z5Kau0vX76sKBZSh0PV1Pj6+mL//v24cOFCpbPqEi4uLnBxcQEAmEwmXL9+HSaTCQ899BB+/PFHq/WHDh2KgIAAu8Rdlueee878f19fX7Rq1QqnT5/GsGHDzOWtWrWCr68vfv31V3PZhg0b0L17d9SvX99ilFF0dDTefvtt7NmzByNHjizzuK+++qpFDUFZSvvFeTd/f38MGzYM69atw/3334/Bgwfj/PnzmDRpElxdXVFUVIRbt25VeJy7TZw4ET/99BO+/vpr1Kv3v7ffjRs3AAAPP/ww/vnPfwK48xp5eHggISEBKSkpiI6OBgCsXbsWR48excaNG8s9li3X8dq1axYjskriuXeUl4eHh/lX+a1bt6DT6Uo9tpubm8W12blzJz7//HPs37+/3Jgr6+rVq9ixYwfmzZuH3Nxc5Obmmp/r27cvZs+ejfPnz6Nhw4bm8vHjx1vUanXv3h2LFy/Gb7/9hsjISHz//fcoLCzEpEmTLNabPHkyFixYoDjmUaNGwdXV1fy4c+fO+PjjjzFmzBiL9Tp37oylS5fi9u3bqFevHr744guYTCYMGzbM4vUIDg5GixYtsHPnTrz++utlHrddu3bYvn17pWIMDg4u9/kJEybgq6++wvDhwzF//nx4enpi+fLlOHjwIADY/HnYs2cP5s6di2HDhqF3797m8pL3X25uLg4dOoSwsDAAQO/evdG8eXMsXLjQ/Dn5/fffMWvWLMycObPc7zRbruPNmzdx8+ZNq33cuHHDYlsXFxfz90jJuev1eqvt3NzczOvo9XoUFhZiypQpeOGFFxAREVGJK1U5MTExeP755zFv3jxs3LgRbm5uWLVqlU37KKmt2rJlC5599lmba6JJXQ6V1CxcuBBxcXEICwtDx44dMWDAAIwaNQpNmzat1Pbr1q3D3/72N5w4cQJFRUXm8pJmnbuVVmZPJcOf72YwGNCoUSOr5hKDwYBr166ZH586dQpHjhwp8wvq0qVL5R47IiLCbl8Uq1atwq1bt/DKK6/glVdeAQD86U9/QrNmzfDFF1/YNATzr3/9K1avXo0333wTAwYMsHiupDlmxIgRFuVPP/00EhISsHfvXkRHR8NoNCIhIQHTpk0zf9GXxZbr+OCDD+K3336zWufebWfPno05c+aYY767WeZu+fn55nO6ffs2XnzxRTzzzDN4+OGHy425sk6fPg0RwcyZMzFz5sxS17l06ZJFUtO4cWOL50v+GJW890rO/97mk4CAgAoT4Mq49/glTZ/3vo4GgwEmkwk5OTlo0KABTp06BREps1nn7kSpNPXr1zcnxEr1798fy5Ytw/Tp09GhQwcAQPPmzTF//ny8+uqrNn0eTpw4gcGDB6NNmzZ4//33LZ4ree907drV4vo0btwY3bp1w969e81lM2bMgJ+fHyZNmlTu8Wy5jgsXLsTcuXOt1pk0aZLFcZo0aWIe1l8Sc0mz9d1KmsNL1lm8eDGuXLlS6jGUWrRoEbZs2YLDhw9j/fr1CAwMtGn74cOH4/3338dzzz2H6dOno0+fPhgyZAieeOIJJji1gEMlNcOGDUP37t2xadMmfPfdd/jrX/+Kv/zlL/jiiy/Qv3//crf95z//idGjR2PQoEGYNm0aAgMD4eLigsTERJw5c8ZqfXv0aShPSa1RZctFxPx/k8mExx57DK+++mqp67Zs2bLcY+fk5FTqF6NOp7Pqq3Ivg8GALVu2ICMjA2fPnkWTJk3QpEkTdOnSBQEBAZVug1+7di1ee+01vPDCC5gxY4bV8yU1c0FBQRblJV9IJX94Fy1ahMLCQgwfPtz8ZVpyz45r167h7NmzCA0NhU6ns+k6fvTRRxbXrOT9d+8v/LsT7JCQEBQXF+PSpUsWX5yFhYX4/fffzef0j3/8AydPnsSqVaus7uuSm5uLs2fPIjAw0FwDVBkmkwkA8Morr6Bv376lrtO8eXOLx5V571Wnqn4mTCYTNBoNvv3221LXrSiRKCwsxNWrVysVY0BAQJnxlJg4cSKeffZZHDlyBDqdDu3bt8eaNWsAVPzZLJGZmYmYmBgYDAZ888038Pb2tni+rM8DcOczUdKf69SpU3jvvfewZMkSXLhwwbxOfn4+ioqKcPbsWfj4+MDPz8+m6zhq1Ch069bN4vnHHnsM06ZNQ0xMjLns7u/RkJAQACj1njsXL16En58f9Ho9cnJy8NZbb+HPf/4zjEYjjEYjgDu1QCKCs2fPwsPDw+ZkpMShQ4fMP1iOHj1q9UOpIu7u7tizZw927tyJr7/+Gtu2bcOnn36K3r1747vvvqvw/UEqU7Hpq0LZ2dnSsGFD6dq1q7nsD3/4Q6l9W2JjY6Vp06YWfQFERLp06WKxfln9L0REDh48WGa7cVnHLW/0073KGpHQpEkTi2HBEREREhUVZbVeZdmrT01Zrl27JjqdztynpSKbN28WFxcXGTp0aJmjjkru97JmzRqL8jNnzggAc3t/Zc7t0KFDIqLsOlamT83WrVtLHf30n//8x6JvwezZsyuMedOmTTbFl52dLQAqNSKj5Fz++9//WpSX9DUp6Rezfv16ASDbtm2zWO/SpUtWfWo2btxoc5+aDRs2VCqukutV0qdh4cKFAkBOnjxZ4bmWxp59asry5JNPiru7u7l/UnmuXLkirVu3lsDAwDJHSxmNRnF1dZXu3btbPde9e3dzn6nKnNtLL70kIsqvY1nfjXcLCAgoc/RT7969ReR/35nlLbGxsVWK8caNG9KsWTNp06aNjB8/XlxcXOTAgQNV2tfd5s+fLwBk+/btivdF1cthamqKi4tx48YNi9E4gYGBCA0NtajO9PT0LHUkQ0n2LCLm5p39+/cjNTXVqtq7LJ6engCA69evl/pcZUdQKDVs2DDMmTMH//rXv6x+hV+/fh1eXl4W/VHuZa8+NWVJSEjA7du3MWXKFIvyEydOwMPDw+J679mzB0899RR69OiBjz76qMzq29jYWLz00ktITk7G6NGjzeuVVMs/9thjAIAXX3wRgwYNstj20qVLeP755zF69GjExsaamxaVXseK9O7dG35+flixYoVFc9qKFSvg4eGBgQMHAgCeeuopq9F3ADB48GAMGDAA48aNQ+fOnW06dmBgIHr27IlVq1Zh0qRJ5l/JJS5fvmxzn7Ho6Gi4urpi2bJliImJMX+OShslVt5nxd6GDBmChIQEzJ07F//85z8tmm9FBFevXkWDBg3K3N6efWpKs3fvXnzxxReYMGGCxffXxYsXkZOTg2bNmpmbdvLy8jBgwACcP38eO3fuLLMpyNvbGwMGDMDWrVtx4sQJtG7dGsCdu/Xu3bsXzz//PACgTZs22LRpk9X2M2bMQG5uLt599100a9YMgPLrWBlDhw7FunXrkJmZaW42S0lJwS+//GL+vggMDCw15qVLlyI1NRUff/yx1fu5sl577TVkZGRg3759aNWqFVJSUhAXF4dDhw6V2tenNFevXrWqwS75/N79t6i077uMjAzcvHnT/HoBd/rlXblyBY0bN7apNpaqSOWkyuzatWvi6ekpcXFx8s4778h7770nw4YNEwDyt7/9zbxeya+NKVOmyPr16833Ovnggw8EgPzxj3+UVatWyfTp08XX11ceeOCBStfUFBYWiq+vr7Rq1Uref/99+fjjj+XXX38t97jVUVOTl5cnHTp0kHr16slzzz0nK1askEWLFpn3W5O98hMTE2XkyJGydOlSWb58ucTExJhHF9wL99T+nD17VgwGg7i7u0tSUpJ8+OGHFsu9IzhKbkj32GOPSVJSkowfP140Gk2FNUJlvaZKrmNlampERJKSkgSAPPHEE7J69WoZNWqURc1SeVDG6KeS+yJV5Pjx41K/fn1p0KCBTJ8+Xd577z158803ZcCAARIZGWl1LhXV1IiIJCQkCAAZMGCA/P3vf5exY8dKaGio+Pv7W9TUXLx4UVxcXOSRRx6RtWvXyscffyzZ2dnm+O1ZUyNy530IQLp06SILFy6UFStWyKuvviotWrSo0TtJnz17Vjp16iRvvfWWvP/++zJlyhRxd3eXBx980GpEVEmt4t21P7GxsQJAxowZY/V5uLe27vjx4+Ll5SUhISGSmJgoiYmJEhISIgEBAXLu3Lly4yzru0bJdbz3e640GRkZ0qBBA2nWrJksXbpUFixYIPXr17capVeasr43S94nFR07JSVFNBqNzJkzx1y2Z88e0Wq1Mm3atHK3vdtLL70kDz74oMyYMUNWr14t8+fPl4YNG0qjRo0sauLu/b4TKf2zW/J+rms3qlSLwyQ1BQUFMm3aNGnXrp14e3uLp6entGvXTpYvX26x3o0bN+Tpp58WX19fAf53EzyTySQLFiyQJk2aiF6vlwcffFC2bt0qcXFxlU5qRES2bNkiERERUq9ePYsPUlnHrY6kRuTOrcMTEhKkefPmotPpxN/fX7p06SKLFi2yGBZb3bZu3SqdOnUSb29v8fDwkEceeUQ+++yzUte990NeUdX43cPYRe68hsuWLZOWLVuKq6urhIWFyYwZMyo83/Je06pex8omNSIi7733nrRq1cp8c7/FixdbNYOWpqykpmPHjhIcHFypY585c0ZGjRolwcHB4urqKg0bNpQ//OEPsnHjRqtzqUxSU1xcLHPnzpWQkJAyb75XYvXq1dK0aVNxcXGx2E91JDUid6aX6Natm/l2+q1bt5b4+PgqN6dUxdWrVyU2NlaCg4NFp9NJeHi4vPbaa1YJjUjpSU2TJk3K/DyU1rydlpYm0dHR4unpKd7e3hIbG1upG/yVd/O9ql7HyiQWIiLHjh2TmJgY8fDwEF9fXxk5cqRkZWVVuF1Z35vLli0rtVn0bkajUZo0aSIdOnSwuFWEyJ1bF2i1WklNTa0wBpE7yVFsbKyEhoaKTqeT0NBQGTFihNV1Z1LjmDQiNdRLkIgqlJubCz8/PyxZsgTx8fFqh0OkumHDhuHs2bM4cOCA2qFQLeAwfWqI6E4fpIYNG2LcuHFqh0KkOhHBrl27zPfkIaoIa2qIiKhOKJmipSzu7u7lTh1Djo9JDRER1QllzRNXIi4uDmvXrq2ZYKhasPmJiIjqhIqG9ts6PQ85HtbUEBERkVPgRBZERETkFJy++clkMuHChQvw9vausD2ViIjqNhFBbm4uQkNDq3UCy/z8/DInxLWFTqczz4JOdSCpuXDhQoWzORMREd0tMzMTjRo1qpZ95+fno4G7F26iWPG+goODkZ6ezsTm/3P6pKZk9tvTp05ZzYRbFnGSGh0Nu0sRUR1n6/d5bm4uWjRvXum/F1VRWFiImyjGSDSETkEvkEKY8FHWeRQWFjKp+f+cPqkpaXLy9vaGj49PpbZhUkNE5Byq+n1eE90V3KGFTlP1pMaFX/FWnD6pISIickQuGg1cFCRPLtDcmTmMzJjUEBERqUCrAVwUVAhpASY19+CQbiIiInIKrKkhIiJSgV2an8gCkxoiIiIVuChsfnKxXyhOg81PRERE5BRYU0NERKQCNj/ZH5MaIiIiFbD5yf7Y/EREREROoc7U1IhGU+vvFOwMdwiuymvgDOdNROqw9fujJr9v2Pxkf6rW1KxYsQKRkZHw8fGBj48PoqKi8O2335qfz8/PR3x8PBo0aAAvLy8MHToU2dnZKkZMRERkHxrc+SNc1YUpjTVVk5pGjRrh7bffRlpaGg4ePIjevXsjNjYWx48fBwBMmTIFX331FTZs2IDdu3fjwoULGDJkiJohExERkYNStfnp8ccft3g8f/58rFixAvv27UOjRo2wZs0arF+/Hr179wYAJCcn4/7778e+ffvwyCOPqBEyERGRXbD5yf4cpk9NcXExNmzYgLy8PERFRSEtLQ1FRUWIjo42r9O6dWs0btwYqampZSY1BQUFKCgoMD82Go3VHjsREZGtOPrJ/lQf/XT06FF4eXlBr9fjhRdewKZNmxAREYGsrCzodDr4+vparB8UFISsrKwy95eYmAiDwWBewsLCqvkMiIiIbHcnqdEoWNQ+A8ejelLTqlUrHD58GPv378eECRMQFxeHn376qcr7S0hIQE5OjnnJzMy0Y7RERETkqFRvftLpdGjevDkAoGPHjvjvf/+Ld999F8OHD0dhYSGuX79uUVuTnZ2N4ODgMven1+uh1+urO2wiIiJF2Pxkf6rX1NzLZDKhoKAAHTt2hKurK1JSUszPnTx5EhkZGYiKilIxQiIiIuWUNT0p62TsrFStqUlISED//v3RuHFj5ObmYv369di1axf+9a9/wWAwYOzYsZg6dSr8/Pzg4+ODSZMmISoqiiOfiIiIyIqqSc2lS5cwatQoXLx4EQaDAZGRkfjXv/6Fxx57DACwePFiaLVaDB06FAUFBejbty+WL1+uZshERER2oVXY/ORwTS0OQCPi3PegNxqNMBgMyMrOho+PT6W24W35K68mpp7g60FENcVoNCIoOBg5OTmV/ptRlWMYDAb8zac53DVV7xlzS4rxsvF0tcZa2zDRIyIiIqeg+ugnIiKiuoijn+yPSQ0REZEKmNTYH5ufiIiIyCmwpoaIiEgFnNDS/pjUEBERqcAFCpufODDUCpMaIiIiFWgV1tRoeUdhK+xTQ0RERE6BNTVEREQqUDz6iRU1VpjUEBERqUBxR2E2P1lh8xMRERE5hTpTU6MR4RxCFaiJeZyqorrjqrPvCzHZvo3Gtt9BVXnt6uzrQXUOm5/sr84kNURERI6EzU/2x+YnIiIicgpMaoiIiFSg1WgUL7ZITEzEww8/DG9vbwQGBmLQoEE4efKkxTr5+fmIj49HgwYN4OXlhaFDhyI7O9uep12tmNQQERGpQOOiUbzYYvfu3YiPj8e+ffuwfft2FBUVISYmBnl5eeZ1pkyZgq+++gobNmzA7t27ceHCBQwZMsTep15t2KeGiIioDti2bZvF47Vr1yIwMBBpaWno0aMHcnJysGbNGqxfvx69e/cGACQnJ+P+++/Hvn378Mgjj6gRtk2Y1BAREalA66KBVqt8mgSj0WhRrtfrodfrK9w+JycHAODn5wcASEtLQ1FREaKjo83rtG7dGo0bN0ZqamqtSGrY/ERERKQGFy00Cha43PkTHhYWBoPBYF4SExMrPLTJZMLkyZPRtWtXtGnTBgCQlZUFnU4HX19fi3WDgoKQlZVl99OvDqypISIiUoFGa3u/GIvtcWfbzMxM+Pj4mMsrU0sTHx+PY8eO4Ycffqjy8R0RkxoiIqJazMfHxyKpqcjEiROxdetW7NmzB40aNTKXBwcHo7CwENevX7eorcnOzkZwcLA9Q642bH4iIiJSgdZFo3ixhYhg4sSJ2LRpE3bs2IHw8HCL5zt27AhXV1ekpKSYy06ePImMjAxERUXZ5ZyrG2tqahFHnMbA1lva19Xb5tfEa2frddJUZZoEW7fR2v4V44jvc1s5w3uWqp9Gq4VGW/W6BVvfZ/Hx8Vi/fj22bNkCb29vcz8Zg8EAd3d3GAwGjB07FlOnToWfnx98fHwwadIkREVF1YpOwgCTGiIiojphxYoVAICePXtalCcnJ2P06NEAgMWLF0Or1WLo0KEoKChA3759sXz58hqOtOqY1BAREamgKk1IFtvD9uaniri5uSEpKQlJSUlVDUtVTGqIiIhUUJW7Altsb2NSUxewozARERE5BdbUEBERqeBOTY2CjsKoQmd/J8ekhoiISAU13aemLmDzExERETkF1tQQERGpQKPRQKNgQkuNiTU192JSQ0REpAKtixZaBX1qtMLGlnsxqSEiIlKB4iHdwpqaezHNIyIiIqfAmhqVOMP8NlXBOXHqmCrML6Wx8beWI86rVSOqMneXrTT83VudWFNjf0xqiIiIVMA+NfbHK0JEREROgTU1REREalDY/AQ2P1lhUkNERKQCrUYDrYL71GjraN/M8rD5iYiIiJwCa2qIiIhUoHHRKpvQ0sR6iXsxqSEiIlKB4gktOU2CFaZ5RERE5BRYU0NERKQCxTffY02NFSY1REREKmCfGvtjUkNmDnkreKo0m6cLqIlb4FfhGE4xhUhNTGFQE2w9D06rYBOtCxT2qbFjME6C70AiIiJyCqomNYmJiXj44Yfh7e2NwMBADBo0CCdPnrRYp2fPntBoNBbLCy+8oFLERERE9qHRahQvZEnVpGb37t2Ij4/Hvn37sH37dhQVFSEmJgZ5eXkW640bNw4XL140LwsXLlQpYiIiIvvQarXmSS2rtGjZ2HIvVfvUbNu2zeLx2rVrERgYiLS0NPTo0cNc7uHhgeDg4JoOj4iIiGoRh0rzcnJyAAB+fn4W5R999BH8/f3Rpk0bJCQk4ObNm2Xuo6CgAEaj0WIhIiJyNCVDupUsZMlhRj+ZTCZMnjwZXbt2RZs2bczlTz/9NJo0aYLQ0FAcOXIEr732Gk6ePIkvvvii1P0kJiZi7ty5NRU2ERFRlSge0q1gW2flMElNfHw8jh07hh9++MGifPz48eb/t23bFiEhIejTpw/OnDmDZs2aWe0nISEBU6dONT82Go0ICwurvsCJiIjIIThEUjNx4kRs3boVe/bsQaNGjcpdt3PnzgCA06dPl5rU6PV66PX6aomTiIjIXjRaLTQKOvsq2dZZqZrUiAgmTZqETZs2YdeuXQgPD69wm8OHDwMAQkJCqjk6IiKi6lMyiknJ9mRJ1aQmPj4e69evx5YtW+Dt7Y2srCwAgMFggLu7O86cOYP169djwIABaNCgAY4cOYIpU6agR48eiIyMVDN0IiIicjCqJjUrVqwAcOcGe3dLTk7G6NGjodPp8P3332PJkiXIy8tDWFgYhg4dihkzZqgQLRERkR0p7CgM1tRYUb35qTxhYWHYvXt3DUXzP44694ytczM56nnURVWZV6vaX78amKenzr4HnWUOJGeZw8pBabQKRz+xT40Vh+goTEREVNewo7D98YoQERGRU2BNDRERkQru3HzPRcH2xXaMxjkwqSEiIlIB7yhsf7wiRERE5BRYU0NERKQCrVYLrYLOvkq2dVZMaoiIiFTA5if74xUhIiIip8CaGiIiIhWwpsb+mNQQERGpQKNRePM9Z7lztR3xihAREZFTYE2NE6vKfENUdzjLvEx8n1cj1gRUKzY/2R+TGiIiIhUwqbE/JjVEREQq0LpooVWQmCjZ1lnxihAREZFTYE0NERGRCjRajbLRT1rn6BdnT0xqiIiIVMA+NfbHK0JEREROgTU1REREKmBNjf0xqSEiIlIB7yhsf7wiRERE5BRYU0NERKQCjYsLtC4uirYnS0xqSlGV267best53tq9dquJKQac4S1SlctU3Z8NR50egt8JdQ/71NgfrwgRERE5BdbUEBERqYA1NfbHpIaIiEgFGq3C0U8KtnVWTGqIiIhUwJoa++MVISIiqgP27NmDxx9/HKGhodBoNNi8ebPF8yKCWbNmISQkBO7u7oiOjsapU6fUCbaKmNQQERGpQKPVmGtrqrTYOKFlXl4e2rVrh6SkpFKfX7hwIZYuXYqVK1di//798PT0RN++fZGfn2+P060RbH4iIiJSQU33qenfvz/69+9f6nMigiVLlmDGjBmIjY0FAPzjH/9AUFAQNm/ejKeeeqrKcdYk1tQQERHVcenp6cjKykJ0dLS5zGAwoHPnzkhNTVUxMtuwpoaIiEgFGq0LNFoFdxT+/9sajUaLcr1eD71eb9O+srKyAABBQUEW5UFBQebnagPW1BAREalB66J8ARAWFgaDwWBeEhMTVT4x9bCmhoiIqBbLzMyEj4+P+bGttTQAEBwcDADIzs5GSEiIuTw7Oxvt27dXHGNNqTNJjWg0lZ7zpSpzsGhMxTatf7sKlWT1HHPKGodj69w+VZpyxwGn6amJKY1s/mw44nXiHEt1is3fBzU5N5hWe2dRsj0AHx8fi6SmKsLDwxEcHIyUlBRzEmM0GrF//35MmDBB0b5rUp1JaoiIiByJxsVF0Uzbtm5748YNnD592vw4PT0dhw8fhp+fHxo3bozJkyfjrbfeQosWLRAeHo6ZM2ciNDQUgwYNqnKMNY1JDRERUR1w8OBB9OrVy/x46tSpAIC4uDisXbsWr776KvLy8jB+/Hhcv34d3bp1w7Zt2+Dm5qZWyDZjUkNERKSGuzr7Vnl7G/Ts2RNSTvOrRqPBvHnzMG/evKrHpDImNURERGrQahUmNRzAfC8mNURERCrgLN32xytCREREToE1NURERGrQKOxTo1GwrZNiUkNERKSGGu4oXBew+YmIiIicAmtqiIiIVMCOwvZXZ5IakcrfDr9KN8nW2PbmcpYLX6O3FK+kYpPj3QZfW4XrpLV1jgHHO20i1dk6LUaNTqPB5ie7Y5pHRERETkHVpCYxMREPP/wwvL29ERgYiEGDBuHkyZMW6+Tn5yM+Ph4NGjSAl5cXhg4diuzsbJUiJiIispOSm+9VeWG9xL1UvSK7d+9GfHw89u3bh+3bt6OoqAgxMTHIy8szrzNlyhR89dVX2LBhA3bv3o0LFy5gyJAhKkZNRESkXMmElkoWsqRq145t27ZZPF67di0CAwORlpaGHj16ICcnB2vWrMH69evRu3dvAEBycjLuv/9+7Nu3D4888ogaYRMREZEDcqi6q5ycHACAn58fACAtLQ1FRUWIjo42r9O6dWs0btwYqamppe6joKAARqPRYiEiInI4Wq3yhSw4zBUxmUyYPHkyunbtijZt2gAAsrKyoNPp4Ovra7FuUFAQsrKySt1PYmIiDAaDeQkLC6vu0ImIiGynqD+NwpFTTsphkpr4+HgcO3YMn3zyiaL9JCQkICcnx7xkZmbaKUIiIiL70WhdFC9kySFulzJx4kRs3boVe/bsQaNGjczlwcHBKCwsxPXr1y1qa7KzsxEcHFzqvvR6PfR6fXWHTERERA5G1ZoaEcHEiROxadMm7NixA+Hh4RbPd+zYEa6urkhJSTGXnTx5EhkZGYiKiqrpcImIiOxHo7A/jY03fa0LVK2piY+Px/r167FlyxZ4e3ub+8kYDAa4u7vDYDBg7NixmDp1Kvz8/ODj44NJkyYhKiqKI5+IiKhWU9qExOYna6omNStWrAAA9OzZ06I8OTkZo0ePBgAsXrwYWq0WQ4cORUFBAfr27Yvly5fXcKRERETk6FRNaqQSc2y4ubkhKSkJSUlJio6lhVR6Lp2qzGdUo/OFOJCaOO8iGw9xq8hk0/r6erZX4bpqbXuP2DyPExE5v5I7CivZniw4REdhIiKiOkfpvWaY1FjhFSEiIiKnwJoaIiIiFSidv4lzP1ljUkNERKQGpXcF5ugnK0xqiIiI1MCkxu7Yp4aIiIicAmtqiIiIVKDRaqFRMIJJybbOikkNERGRGjQKm580bH66F9M8IiIicgqsqSEiIlKDRqNsUsoq3P3e2TGpKUVdnfKgJphg+4ewsNi2aQ/cXW37knCt0vcC3yNEpJBG4UzbnKXbCq8IEREROQXW1BAREalANFqIgtoWJds6KyY1REREamDzk90xqSEiIqIaYzQaK72uj4+PTftmUkNERKQGjUbZCKZaOvrJ19cXmkrGXlxcbNO+mdQQERGpQau9syjZvhbauXOn+f9nz57F9OnTMXr0aERFRQEAUlNTsW7dOiQmJtq8byY1REREKqirHYUfffRR8//nzZuHd955ByNGjDCX/fGPf0Tbtm3x3nvvIS4uzqZ9184rQkRERLVeamoqHnroIavyhx56CAcOHLB5f0xqiIiI1FAy+knJUsuFhYVh9erVVuXvv/8+wsLCbN4fm5+IiIjUwCHdWLx4MYYOHYpvv/0WnTt3BgAcOHAAp06dwueff27z/mr/FSEiIqJaacCAAfjll1/w+OOP4+rVq7h69Soef/xx/PLLLxgwYIDN+2NNTSmklg6Tu1eNzGElts3LpK3CLwvPeo73emiKC21cv8jmY5h0njZvQ0S1CGtqANxpglqwYIFd9sWkhoiISAWi0Sgc/eR4P/gq48iRI2jTpg20Wi2OHDlS7rqRkZE27ZtJDREREdWY9u3bIysrC4GBgWjfvj00Gg2klJYFjUbDm+8RERHVCnW0+Sk9PR0BAQHm/9sTkxoiIiI11NFpEpo0aQIAKCoqwty5czFz5kyEh4fbZd+1M80jIiKiWs3V1bVKw7bLw6SGiIhIDbz5HgYNGoTNmzfbbX9sfiIiIlJBXZ376W4tWrTAvHnz8J///AcdO3aEp6flrSxefPFFm/bHpIaIiEgNGoWzdDtBUrNmzRr4+voiLS0NaWlpFs9pNBomNURERFQ7cPQTERGRM6ijQ7qrU51JakzQwITKDX+riUFyNTKFQQ0QrYtN62tvF9h+DBedzdtUN1tjcsRzICKVManB1KlTSy3XaDRwc3ND8+bNERsbCz8/v0rtr84kNURERORYDh06hB9//BHFxcVo1aoVAOCXX36Bi4sLWrdujeXLl+Pll1/GDz/8gIiIiAr3V/vTPCIiotqIQ7oRGxuL6OhoXLhwwdxZ+Ny5c3jssccwYsQInD9/Hj169MCUKVMqtT+NlDbhghMxGo0wGAy4mJUNHx+fSm1TEzdpdJrmJxsvlrM0PxGRczIajQgKDkZOTk6l/2ZU5RgGgwGXMtMVHcNoNCIwLLxaY61uDRs2xPbt261qYY4fP46YmBicP38eP/74I2JiYnDlypUK91f70zwiIiKqlXJycnDp0iWr8suXL8NoNAIAfH19UVhYWKn9MakhIiJSA5ufEBsbizFjxmDTpk04d+4czp07h02bNmHs2LEYNGgQAODAgQNo2bJlpfbHjsJERERqqKMTWt5t1apVmDJlCp566incvn0bAFCvXj3ExcVh8eLFAIDWrVvj/fffr9T+mNQQERGRKry8vLB69WosXrwYv/76KwCgadOm8PLyMq/Tvn17nDt3DiaTCdoK7sBc++uuiIiIaiOVmp+SkpJw3333wc3NDZ07d8aBAwfsfGK28/LyQmRkJCIjIy0SmhIRERE4e/Zshfux+YrExcVhz549tm5GREREdymZ0FLJYqtPP/0UU6dOxezZs/Hjjz+iXbt26Nu3b6mddR1JZQdq23xFcnJyEB0djRYtWmDBggU4f/68zcERERHVeSrU1LzzzjsYN24cnn32WURERGDlypXw8PDABx98UA0nWPNsviKbN2/G+fPnMWHCBHz66ae477770L9/f2zcuBFFRUXVESMRERGVwWg0WiwFBaXfD6ywsBBpaWmIjo42l2m1WkRHRyM1NbWmwq1WVeooHBAQgKlTp2Lq1Kn48ccfkZycjGeeeQZeXl7405/+hD//+c9o0aKFvWNVRAuBFpWrvpIqzP7kLDfTs5Wt5y1a299yLrnZNq2vOX/CpvUndnjBpvUB4N0bR21avyrnTUTOTTQam29geu/2ABAWFmZRPnv2bMyZM8dq/StXrqC4uBhBQUEW5UFBQThxwrbvTUelqKPwxYsXsX37dmzfvh0uLi4YMGAAjh49ioiICPNQrPLs2bMHjz/+OEJDQ6HRaLB582aL50ePHg2NRmOx9OvXT0nIREREDkFE+QIAmZmZyMnJMS8JCQnqnlg10FQy+bP552NRURG+/PJLJCcn47vvvkNkZCQmT56Mp59+2nyb5k2bNmHMmDEVztWQl5eHdu3aYcyYMRgyZEip6/Tr1w/Jycnmx3q93taQiYiInJaPj0+lpknw9/eHi4sLsrMta7+zs7MRHBxcXeHZRWU7Ctuc1ISEhMBkMmHEiBE4cOAA2rdvb7VOr1694OvrW+G++vfvj/79+5e7jl6vd/iLTUREZCuTCEwKui7Yuq1Op0PHjh2RkpJivluvyWRCSkoKJk6cWOU4asJPP/2E0NDQCtezuflp8eLFuHDhApKSkkpNaIA78zSkp6fbuutS7dq1C4GBgWjVqhUmTJiA33//3S77JSIiUpPYYbHV1KlTsXr1aqxbtw4///wzJkyYgLy8PDz77LOKz8cW9w4hP3z4MOLi4tC1a1c88cQT2LVrl8XzYWFhcHFxqXC/NtfUPPPMM7ZuUmX9+vXDkCFDEB4ejjNnzuD1119H//79kZqaWubJFRQUWPT8LpkQi4iIqK4bPnw4Ll++jFmzZiErKwvt27fHtm3brDoPV7eQkBBcvHgRgYGB2Lt3L3r27IkuXbqga9euOHz4MB577DGkpKSgR48eNu3XoYdkPPXUU+b/t23bFpGRkWjWrBl27dqFPn36lLpNYmIi5s6dW1MhEhERVYlJ7ixKtq+KiRMnqt7cdHcfmTlz5uCZZ57BmjVrzGWTJ0/G3LlzkZKSYtN+a9U0CU2bNoW/vz9Onz5d5joJCQkWvcAzMzNrMEIiIqLKERHFizM4duwYxo0bZ1E2btw4HDlyxOZ9OXRNzb3OnTuH33//HSEhIWWuo9frOUKKiIjIweXm5sLNzQ1ubm5Wf7fd3Nxw8+ZNm/epalJz48YNi1qX9PR0HD58GH5+fvDz88PcuXMxdOhQBAcH48yZM3j11VfRvHlz9O3bV8WoiYiIlFOr+clRtGzZEsCdGquDBw/iwQcfND93/PjxSo12upeqSc3BgwfRq1cv8+OpU6cCuDNp5ooVK3DkyBGsW7cO169fR2hoKGJiYvDmm2+yJoaIiJxCLc9Lqmznzp0Wj+9tgUlPT8f48eNt3q9GnKVRrgxGoxEGgwHZWVmVujlRjRGT7dvYOHmZNj/HpvWlnptN69/ZpvoTTG1Brk3r2zqtQpF/c5vWdxpVeQ/WhCpM0kdkL0ajEUHBwcjJyam2vxklf5fOZF6Et4Jj5BqNaBYWUq2x1jb89iAiIiKnwKSGiIhIBXV99NPy5csRHR2NYcOGWQ3dvnLlCpo2bWrzPpnUEBERqcBkh6W2Wrp0KaZNm4bWrVtDr9djwIABSExMND9fXFyM3377zeb91qoh3URERFT7rVq1CqtXr8bTTz8NAJgwYQIGDRqEW7duYd68eVXeL5MaIiIiFYjcWZRsX1ulp6ejS5cu5sddunTBjh07EB0djaKiIkyePLlK+2VSQ0REpIK6fJ8af39/ZGZm4r777jOXtWnTBjt27EDv3r1x4cKFKu2XfWqIiIioRnXr1g1ffPGFVXlERARSUlLw7bffVmm/rKkhIiJSgdIRTLV59NP06dORlpZW6nMPPPAAduzYgc8//9zm/TKpISIiUoHSEUy1efRTZGQkIiMjy3y+TZs2aNOmjc37ZVJDREREqrl+/ToOHDiAS5cuwWT6X6qm0WjwzDPP2LQvJjVEREQqECgc/WS3SNTz1VdfYeTIkbhx4wZ8fHyg0WjMzzGpqU2qMr+NjXP1mNwMth/DRnm3q/9j5an3tml9cXWvpkhqltz14a4MTQ20r2tsfA9KDbzPbca5pchBmERgUvC5VbKto3j55ZcxZswYLFiwAB4eHor3x6SGiIhIBQJltS21P6UBzp8/jxdffNEuCQ3AId1ERESkkr59++LgwYN22x9raoiIiFRQV2++9+WXX5r/P3DgQEybNg0//fQT2rZtC1dXV4t1//jHP9q0byY1REREalA4TUJtbX8aNGiQVVlp8z1pNBoUFxfbtG8mNURERFRj7h62bW9MaoiIiFRggsCkoLpFybaOorwZuTUaDWbOnGnT/pjUEBERqaAuz9JdYtOmTRaPi4qKkJ6ejnr16qFZs2ZMaoiIiKh2OHTokFWZ0WjE6NGjMXjwYJv3xyHdREREKigZ/aRkcUY+Pj6YO3euzbU0AGtqiIiIVMHmp7Ll5OQgJyfH5u2Y1BAREZEqli5davFYRHDx4kV8+OGH6N+/v837Y1JTmqrMPVMT88nYeAwTbJs7SCu23Q8AADzrOV4Lpmhr4G1d3fMTARC42LaBrXNFVaH12eYfhpxniahMHP0ELF682OKxVqtFQEAA4uLikJCQYPP+mNQQERGpgM1PQHp6ul33x6SGiIhIBZyl2/5YN0xEREROgTU1REREKig23VmUbE+WmNQQERGpgM1P9sfmJyIiInIKrKkhIiJSgUkExaypsSsmNURERCq4M9WBkqTGjsE4CTY/ERERkVNgTQ0REZEKOPrJ/pjUlMZJbu2utfEW2qK18bb8ADTO0KZbhSkPNKbbNm5g+3tKa+s2NfFSOMlng8gRcPST/fEbioiIiJwCa2qIiIhUUKxw9JOSbZ0VkxoiIiIVmKBsBBO71FhjUkNERKSCYpOgWEFWo2RbZ8U+NUREROQUWFNDRESkAlE4+knYp8YKkxoiIiIVFMudRcn2ZInNT0REROQUWFNDRESkAt58z/6Y1BAREamAo5/sj81PRERE5BRYU0NmTjGPEwDRaGxaX1OF3F60/OgQkTJsfrI/VWtq9uzZg8cffxyhoaHQaDTYvHmzxfMiglmzZiEkJATu7u6Ijo7GqVOn1AmWiIjIjkpGPylZyJKqSU1eXh7atWuHpKSkUp9fuHAhli5dipUrV2L//v3w9PRE3759kZ+fX8OREhERkaNTtQ69f//+6N+/f6nPiQiWLFmCGTNmIDY2FgDwj3/8A0FBQdi8eTOeeuqpmgyViIjIrtj8ZH8O21E4PT0dWVlZiI6ONpcZDAZ07twZqampZW5XUFAAo9FosRARETkak0kUL2TJYZOarKwsAEBQUJBFeVBQkPm50iQmJsJgMJiXsLCwao2TiIioKkwK+9Mwp7HmsElNVSUkJCAnJ8e8ZGZmqh0SERER1QCHHZcaHBwMAMjOzkZISIi5PDs7G+3bty9zO71eD71eX93hERERKcI+NfbnsDU14eHhCA4ORkpKirnMaDRi//79iIqKUjEyIiIi5YpFFC9kSdWk5saNGzh8+DAOHz4M4E7n4MOHDyMjIwMajQaTJ0/GW2+9hS+//BJHjx7FqFGjEBoaikGDBqkZNhERkVObP38+unTpAg8PD/j6+pa6TkZGBgYOHAgPDw8EBgZi2rRpuH37ds0Geg9Vm58OHjyIXr16mR9PnToVABAXF4e1a9fi1VdfRV5eHsaPH4/r16+jW7du2LZtG9zc3NQKmYiIyC6UjmCqztFPhYWFePLJJxEVFYU1a9ZYPV9cXIyBAwciODgYe/fuxcWLFzFq1Ci4urpiwYIF1RZXRTQizl1/ZTQaYTAYkJ2VBR8fH7XDqVE2Txfg3G8FIqIKGY1GBAUHIycnp9r+ZpT8XVqcchTunt5V3s+tvFxM6dO2WmNdu3YtJk+ejOvXr1uUf/vtt/jDH/6ACxcumEcpr1y5Eq+99houX74MnU5XLfFUxGH71BAREVHF7r03W0FBQbUfMzU1FW3btrW47Urfvn1hNBpx/Pjxaj9+WZjUEBERqaBk9JOSBQDCwsIs7s+WmJhY7bFnZWWVeh+5kufU4rBDuomIiJyZ0hFMJdtmZmZaND+VdVuT6dOn4y9/+Uu5+/z555/RunXrKsekNiY1REREtZiPj0+l+tS8/PLLGD16dLnrNG3atFLHDA4OxoEDByzKsrOzzc+phUkNERGRCkwmQXENjn4KCAhAQEBAlY93t6ioKMyfPx+XLl1CYGAgAGD79u3w8fFBRESEXY5RFUxqiIiIVFCsMKlRsm1FMjIycPXqVWRkZKC4uNh8P7nmzZvDy8sLMTExiIiIwDPPPIOFCxciKysLM2bMQHx8vKp39WdSQ0REpAJHTmpmzZqFdevWmR8/+OCDAICdO3eiZ8+ecHFxwdatWzFhwgRERUXB09MTcXFxmDdvXrXFVBlMaoiIiMjC2rVrsXbt2nLXadKkCb755puaCaiSmNQQERGpoNikrLal2GTHYJwEkxoiIiIVOHLzU23Fm+8RERGRU2BNjRPjXE5ERI6LNTX2x6SGiIhIBTV9n5q6gM1PRERE5BRYU0NERKSCYlHY/MQuBlaY1BAREamAfWrsj81PRERE5BRYU0NERKQC1tTYH5MaIiIiFdw2CVwUJCa3mdRYYVJDRESkAtbU2B/71BAREZFTYE0NERGRCnjzPftjUkNERKSCYhFF95rhfWqssfmJiIiInAJraoiIiFTAjsL2x6SGiIhIBUxq7I/NT0REROQUWFNDRESkAtbU2B+TGiIiIhUUiwnFJpOi7ckSm5+IiIjIKbCmhoiISAW8+Z79MakhIiJSQbFJoGWfGrtiUkNERKSC2yZAo2iWbjsG4yTYp4aIiIicAmtqiIiIVMDmJ/tjUkNERKQCJjX2x+YnIiIicgqsqSEiIlIBa2rsj0kNERGRCnifGvtj8xMRERE5BdbUEBERqaDYJIruU8PmJ2tMaoiIiFQgIhAFiYkIk5p7sfmJiIiInAJraoiIiFRgMomizr7sKGyNSQ0REZEKRERRExKbn6wxqSEiIlKBmBT2qWFNjRX2qSEiIiKn4NBJzZw5c6DRaCyW1q1bqx0WERGRYiV9apQsZMnhm58eeOABfP/99+bH9eo5fMhEREQVEtOdRcn2ZMnhM4R69eohODhY7TCIiIjIwTl08xMAnDp1CqGhoWjatClGjhyJjIyMctcvKCiA0Wi0WIiIiBxNyegnJQtZcuikpnPnzli7di22bduGFStWID09Hd27d0dubm6Z2yQmJsJgMJiXsLCwGoyYiIioctinxv4cOqnp378/nnzySURGRqJv37745ptvcP36dXz22WdlbpOQkICcnBzzkpmZWYMRExERkVocvk/N3Xx9fdGyZUucPn26zHX0ej30en0NRkVERGQ73qfG/hy6puZeN27cwJkzZxASEqJ2KERERMr8/6SmqguY1Fhx6KTmlVdewe7du3H27Fns3bsXgwcPhouLC0aMGKF2aERERORgHLr56dy5cxgxYgR+//13BAQEoFu3bti3bx8CAgLUDo2IiEgRkwg0CkYwmTj6yYpDJzWffPKJ3fYlGg1Eo6nUukreZERERJUhorBPDf9WWXHopIaIiMhZsaOw/Tl0nxoiIiKiymJNDRERkQpMJkCjoLbFxLmfrDCpISIiUoHSqQ7Yp8Yam5+IiIjIKbCmhoiISAViurMo2Z4sMakhIiJSgckkCvvUsPnpXmx+IiIiIqfAmhoiIiIV8D419sekhoiISAVMauyPzU9ERERkdvbsWYwdOxbh4eFwd3dHs2bNMHv2bBQWFlqsd+TIEXTv3h1ubm4ICwvDwoULVYr4f+pMTY1G4cRh9lbZeaju5kjxExGRMo46oeWJEydgMpmwatUqNG/eHMeOHcO4ceOQl5eHRYsWAQCMRiNiYmIQHR2NlStX4ujRoxgzZgx8fX0xfvz4aomrMupMUkNERORIHLX5qV+/fujXr5/5cdOmTXHy5EmsWLHCnNR89NFHKCwsxAcffACdTocHHngAhw8fxjvvvKNqUsPmJyIiIhWUzNJd5aUGa+9zcnLg5+dnfpyamooePXpAp9OZy/r27YuTJ0/i2rVrNRbXvZjUEBER1WJGo9FiKSgosOv+T58+jWXLluH55583l2VlZSEoKMhivZLHWVlZdj2+LZjUEBERqUBMApOCpaT5KSwsDAaDwbwkJiaWerzp06dDo9GUu5w4ccJim/Pnz6Nfv3548sknMW7cuGq/JkqxTw0REZEK7DWhZWZmJnx8fMzler2+1PVffvlljB49utx9Nm3a1Pz/CxcuoFevXujSpQvee+89i/WCg4ORnZ1tUVbyODg4uNLnYG9MaoiIiGoxHx8fi6SmLAEBAQgICKjUPs+fP49evXqhY8eOSE5OhlZr2bATFRWFN954A0VFRXB1dQUAbN++Ha1atUL9+vVtPwk7YfMTERGRChR1ElY4cqo858+fR8+ePdG4cWMsWrQIly9fRlZWlkVfmaeffho6nQ5jx47F8ePH8emnn+Ldd9/F1KlTqyWmymJNDRERkQpMJgEccELL7du34/Tp0zh9+jQaNWpk8VxJk5fBYMB3332H+Ph4dOzYEf7+/pg1a5aqw7kBJjVERER0l9GjR1fY9wYAIiMj8e9//7v6A7IBkxoiIiIViKkYYipWtD1ZYlKjEk55QERUtzGpsT92FCYiIiKnwJoaIiIiFYjJpLCmxmTHaJwDkxoiIiIVSHExpFhBUqNgW2fFpIaIiEgFIgr71AiTmnuxTw0RERE5BdbUEBERqYCjn+yPSQ0REZEKmNTYH5ufiIiIyCmwpoaIiEgFrKmxPyY1REREKuB9auyPzU9ERETkFFhTQ0REpAKTqRhQUFNjYvOTFSY1REREKmCfGvtj8xMRERE5BdbUEBERqYA1NfbHpIaIiEgNxcUQrYLEhBNaWmFSQ0REpAIRZR2FOaGlNfapISIiIqfAmhoiIiIViMmkrKaGN9+zwqSGiIhIBaLwPjXsKGyNzU9ERETkFFhTQ0REpII7zU9Vb0Ji85M1JjVEREQqYPOT/bH5iYiIiJxCrUhqkpKScN9998HNzQ2dO3fGgQMH1A6JiIhIkZI7CitZyJLDJzWffvoppk6ditmzZ+PHH39Eu3bt0LdvX1y6dEnt0IiIiKrMZCpWvJAlh09q3nnnHYwbNw7PPvssIiIisHLlSnh4eOCDDz5QOzQiIiJyIA7dUbiwsBBpaWlISEgwl2m1WkRHRyM1NbXUbQoKClBQUGB+bDQaqz1OIiIiW0mxCdAo6ChczNFP93LomporV66guLgYQUFBFuVBQUHIysoqdZvExEQYDAbzEhYWVhOhEhER2UREYZ8azv1kxaGTmqpISEhATk6OecnMzFQ7JCIiIivsKGx/Dt385O/vDxcXF2RnZ1uUZ2dnIzg4uNRt9Ho99Hp9TYRHREREDsSha2p0Oh06duyIlJQUc5nJZEJKSgqioqJUjIyIiEgZ1tTYn0PX1ADA1KlTERcXh4ceegidOnXCkiVLkJeXh2effVbt0IiIiKpMTMXKOgozqbHi8EnN8OHDcfnyZcyaNQtZWVlo3749tm3bZtV5uCwiAgDIzc2tzjCJiMgJlPytKPnbUa2Ki6DoKMVF9orEaWikRl459Zw7d44joIiIyCaZmZlo1KhRtew7Pz8f4eHhZY7itUVwcDDS09Ph5uZmh8hqP6dPakwmEy5cuABvb29oNBpzudFoRFhYGDIzM+Hj46NihDWL583zrgt43jzvqhIR5ObmIjQ0FFpt9XU7zc/PR2FhoeL96HQ6JjR3cfjmJ6W0Wm252baPj0+d+vCX4HnXLTzvuoXnrYzBYLBDNOVzc3NjMlINHHr0ExEREVFlMakhIiIip1Bnkxq9Xo/Zs2fXuRv18bx53nUBz5vnTXWT03cUJiIiorqhztbUEBERkXNhUkNEREROgUkNEREROQUmNUREROQU6mxSk5SUhPvuuw9ubm7o3LkzDhw4oHZI1WrOnDnQaDQWS+vWrdUOy+727NmDxx9/HKGhodBoNNi8ebPF8yKCWbNmISQkBO7u7oiOjsapU6fUCdaOKjrv0aNHW73+/fr1UydYO0lMTMTDDz8Mb29vBAYGYtCgQTh58qTFOvn5+YiPj0eDBg3g5eWFoUOHIjs7W6WI7aMy592zZ0+r1/uFF15QKWL7WLFiBSIjI8032IuKisK3335rft4ZX2uyXZ1Maj799FNMnToVs2fPxo8//oh27dqhb9++uHTpktqhVasHHngAFy9eNC8//PCD2iHZXV5eHtq1a4ekpKRSn1+4cCGWLl2KlStXYv/+/fD09ETfvn2Rn59fw5HaV0XnDQD9+vWzeP0//vjjGozQ/nbv3o34+Hjs27cP27dvR1FREWJiYpCXl2deZ8qUKfjqq6+wYcMG7N69GxcuXMCQIUNUjFq5ypw3AIwbN87i9V64cKFKEdtHo0aN8PbbbyMtLQ0HDx5E7969ERsbi+PHjwNwzteaqkDqoE6dOkl8fLz5cXFxsYSGhkpiYqKKUVWv2bNnS7t27dQOo0YBkE2bNpkfm0wmCQ4Olr/+9a/msuvXr4ter5ePP/5YhQirx73nLSISFxcnsbGxqsRTUy5duiQAZPfu3SJy57V1dXWVDRs2mNf5+eefBYCkpqaqFabd3XveIiKPPvqovPTSS+oFVUPq168v77//fp15ralida6mprCwEGlpaYiOjjaXabVaREdHIzU1VcXIqt+pU6cQGhqKpk2bYuTIkcjIyFA7pBqVnp6OrKwsi9feYDCgc+fOTv/aA8CuXbsQGBiIVq1aYcKECfj999/VDsmucnJyAAB+fn4AgLS0NBQVFVm83q1bt0bjxo2d6vW+97xLfPTRR/D390ebNm2QkJCAmzdvqhFetSguLsYnn3yCvLw8REVF1ZnXmirm9BNa3uvKlSsoLi5GUFCQRXlQUBBOnDihUlTVr3Pnzli7di1atWqFixcvYu7cuejevTuOHTsGb29vtcOrEVlZWQBQ6mtf8pyz6tevH4YMGYLw8HCcOXMGr7/+Ovr374/U1FS4uLioHZ5iJpMJkydPRteuXdGmTRsAd15vnU4HX19fi3Wd6fUu7bwB4Omnn0aTJk0QGhqKI0eO4LXXXsPJkyfxxRdfqBitckePHkVUVBTy8/Ph5eWFTZs2ISIiAocPH3b615oqp84lNXVV//79zf+PjIxE586d0aRJE3z22WcYO3asipFRTXjqqafM/2/bti0iIyPRrFkz7Nq1C3369FExMvuIj4/HsWPHnLKfWHnKOu/x48eb/9+2bVuEhISgT58+OHPmDJo1a1bTYdpNq1atcPjwYeTk5GDjxo2Ii4vD7t271Q6LHEida37y9/eHi4uLVa/47OxsBAcHqxRVzfP19UXLli1x+vRptUOpMSWvb11/7QGgadOm8Pf3d4rXf+LEidi6dSt27tyJRo0amcuDg4NRWFiI69evW6zvLK93Weddms6dOwNArX+9dTodmjdvjo4dOyIxMRHt2rXDu+++6/SvNVVenUtqdDodOnbsiJSUFHOZyWRCSkoKoqKiVIysZt24cQNnzpxBSEiI2qHUmPDwcAQHB1u89kajEfv3769Trz0AnDt3Dr///nutfv1FBBMnTsSmTZuwY8cOhIeHWzzfsWNHuLq6WrzeJ0+eREZGRq1+vSs679IcPnwYAGr1610ak8mEgoICp32tqQrU7qmshk8++UT0er2sXbtWfvrpJxk/frz4+vpKVlaW2qFVm5dffll27dol6enp8p///Eeio6PF399fLl26pHZodpWbmyuHDh2SQ4cOCQB555135NChQ/Lbb7+JiMjbb78tvr6+smXLFjly5IjExsZKeHi43Lp1S+XIlSnvvHNzc+WVV16R1NRUSU9Pl++//146dOggLVq0kPz8fLVDr7IJEyaIwWCQXbt2ycWLF83LzZs3zeu88MIL0rhxY9mxY4ccPHhQoqKiJCoqSsWolavovE+fPi3z5s2TgwcPSnp6umzZskWaNm0qPXr0UDlyZaZPny67d++W9PR0OXLkiEyfPl00Go189913IuKcrzXZrk4mNSIiy5Ytk8aNG4tOp5NOnTrJvn371A6pWg0fPlxCQkJEp9NJw4YNZfjw4XL69Gm1w7K7nTt3CgCrJS4uTkTuDOueOXOmBAUFiV6vlz59+sjJkyfVDdoOyjvvmzdvSkxMjAQEBIirq6s0adJExo0bV+uT+NLOF4AkJyeb17l165b8+c9/lvr164uHh4cMHjxYLl68qF7QdlDReWdkZEiPHj3Ez89P9Hq9NG/eXKZNmyY5OTnqBq7QmDFjpEmTJqLT6SQgIED69OljTmhEnPO1JttpRERqrl6IiIiIqHrUuT41RERE5JyY1BAREZFTYFJDREREToFJDRERETkFJjVERETkFJjUEBERkVNgUkNEREROgUkNEREROQUmNUREROQUmNQQERGRU2BSQ1SHXL58GcHBwViwYIG5bO/evdDpdBYzHBMR1Uac+4mojvnmm28waNAg7N27F61atUL79u0RGxuLd955R+3QiIgUYVJDVAfFx8fj+++/x0MPPYSjR4/iv//9L/R6vdphEREpwqSGqA66desW2rRpg8zMTKSlpaFt27Zqh0REpBj71BDVQWfOnMGFCxdgMplw9uxZtcMhIrIL1tQQ1TGFhYXo1KkT2rdvj1atWmHJkiU4evQoAgMD1Q6NiEgRJjVEdcy0adOwceNG/N///R+8vLzw6KOPwmAwYOvWrWqHRkSkCJufiOqQXbt2YcmSJfjwww/h4+MDrVaLDz/8EP/+97+xYsUKtcMjIlKENTVERETkFFhTQ0RERE6BSQ0RERE5BSY1RERE5BSY1BAREZFTYFJDREREToFJDRERETkFJjVERETkFJjUEBERkVNgUkNEREROgUkNEREROQUmNUREROQUmNQQERGRU/h/qlVN0ZNjGlMAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# gridded cloud top temperature\n",
        "ds['ctt_grid'].isel(starttime=0).plot()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "dzBY7eYRfncK",
        "outputId": "fb251f38-3b2d-4371-aace-bd433bd1fe8a"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.QuadMesh at 0x7f0a98d3ce80>"
            ]
          },
          "metadata": {},
          "execution_count": 11
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Now we will preprocess our training and testing data. We need to split our data into a training and testing sets, and standardize."
      ],
      "metadata": {
        "id": "UlzHzBsPhpe1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def random_split(data, label, split=0.3, seed=0):\n",
        "    \"\"\"\n",
        "    Help spliting data randomly for training and testing.\n",
        "    \"\"\"\n",
        "    np.random.seed(0)\n",
        "    da_indx = np.random.permutation(data.shape[0])\n",
        "    data  = data[da_indx.astype(int)]\n",
        "    label = label[da_indx.astype(int)]\n",
        "    init_range = int(data.shape[0] * (1 - 0.3))\n",
        "    return data[:init_range], label[:init_range], data[init_range:], label[init_range:]"
      ],
      "metadata": {
        "id": "BYfUl7q0IJBl"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "features = xr.concat([\n",
        "    ds['dbz_grid'].expand_dims(dim='feature', axis=3),\n",
        "    ds['ctt_grid'].expand_dims(dim='feature', axis=3)],\n",
        "          dim='feature')"
      ],
      "metadata": {
        "id": "QmZSDOrvd50f"
      },
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(features.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rt4Sl3cii7SY",
        "outputId": "37874fa4-eb35-4a44-da37-23f500208608"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(2224, 32, 32, 2)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# grab the maximum of the dbz to use as our labels (regression)\n",
        "labels = ds['dbz_grid'].expand_dims(dim='feature', axis=3).max('x').max('y')"
      ],
      "metadata": {
        "id": "HgsBcfJ2ivWE"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(labels)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kZ57QrQXi3HV",
        "outputId": "93ba6551-8ece-43bf-b441-3968779e586c"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<xarray.DataArray 'dbz_grid' (starttime: 2224, feature: 1)>\n",
            "array([[65.721054],\n",
            "       [58.54062 ],\n",
            "       [58.54062 ],\n",
            "       ...,\n",
            "       [45.41483 ],\n",
            "       [37.786957],\n",
            "       [42.48563 ]], dtype=float32)\n",
            "Coordinates:\n",
            "  * starttime  (starttime) float64 9.264e+04 9.264e+04 ... 9.266e+04 9.266e+04\n",
            "    endtime    (starttime) float64 ...\n",
            "    x_speed    (starttime) int64 ...\n",
            "    y_speed    (starttime) int64 ...\n",
            "Dimensions without coordinates: feature\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_feats, train_label, test_feats, test_label = random_split(features, labels)"
      ],
      "metadata": {
        "id": "1xt_5E0DjOPD"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(train_feats.shape[0], train_label.shape[0], test_feats.shape[0], test_label.shape[0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "q5kQ6CXcjOek",
        "outputId": "6366acef-8999-418e-f0f6-c9d82e23aaba"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1556 1556 668 668\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# normalization of our training data (note each feature individually processed!)\n",
        "train_feats_standardized = xr.concat([\n",
        "    z_score(train_feats.isel(feature=0).expand_dims(dim='feature', axis=3)),\n",
        "    z_score(train_feats.isel(feature=1).expand_dims(dim='feature', axis=3))],\n",
        "                                     dim='feature')\n",
        "\n",
        "train_label_standardized = z_score(train_label)"
      ],
      "metadata": {
        "id": "-UM89woPj8Ep"
      },
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_feats.isel(feature=0,starttime=15).plot()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "1HV4mCLnlIMF",
        "outputId": "8d425663-ace2-49ba-aa92-ef4783520727"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.QuadMesh at 0x7f0a98c6c4c0>"
            ]
          },
          "metadata": {},
          "execution_count": 20
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjUAAAHHCAYAAABHp6kXAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABoIUlEQVR4nO3dd3hUVfoH8O/MJDPpvUMSEnoLJQgbUKTE0FTaT1BQiggLBpbiKsRCUwmiC4hLs4G6qIiCBRsddAkgAaRJDxIgCTWdZJKZ8/uDzciQhOTMTHKTyffzPPeB3HnvPWfunfLOueeeoxJCCBARERHVcmqlK0BERERkC0xqiIiIyC4wqSEiIiK7wKSGiIiI7AKTGiIiIrILTGqIiIjILjCpISIiIrvApIaIiIjsApMaIiIisgtMaqx0/vx5qFQqrF69WumqEFWZBg0aYNSoUZWK7datG7p161al9SEiKkutS2o+/fRTLF68uEr2ffnyZcyePRuHDh2q1nJrus8//xzt27eHk5MT/P39MWbMGFy7dq3C7fLz87F06VLExcUhODgY7u7uaNeuHZYvXw6DwVDmNmfPnsWwYcMQEBAAZ2dnNG7cGC+99FKpOKPRiOXLl6Nt27ZwdnaGr68vevTogd9//93q5yvrgw8+QPPmzeHk5ITGjRvjnXfeqXCbhx56CCqVChMnTqyGGtre8ePHMXv2bJw/f17pqlS7M2fO4P/+7//g7e0NFxcX3H///di+fXultt26dSuefvppNGnSBC4uLoiMjMQzzzyDtLS0MuP1ej3mzZuHZs2awcnJCYGBgejXrx8uXrxYKvbAgQN49NFH4ePjAxcXF7Rq1QpLliyx6rla4o8//kDv3r3h5uYGHx8fPPXUU7h69eo9t1mzZg1UKhXc3NyqqZZkrxyUroCsTz/9FEePHsWUKVNsvu/Lly9jzpw5aNCgAdq2bVupcsPDw3Hr1i04OjravD41wfLly/Hss8+iZ8+eWLhwIS5evIi3334b+/fvx969e+Hk5FTutufOncOkSZPQs2dPTJs2DR4eHvj555/x7LPPYs+ePfjoo4/M4g8dOoRu3bqhXr16eO655+Dr64sLFy4gNTW11L6ffvpprFmzBiNGjMDEiRORl5eHgwcP4sqVKzY/BveycuVKjB8/HoMHD8a0adPwyy+/4B//+Afy8/Mxffr0MrdZv349kpKSqrWetnb8+HHMmTMH3bp1Q4MGDcwe27RpkzKVqgapqamIiYmBRqPB888/D1dXV6xatQpxcXHYunUrunbtes/tp0+fjhs3buCxxx5D48aNce7cOfz73//Gxo0bcejQIQQFBZlii4qK0K9fP+zevRtjx45FVFQUbt68ib179yIrKwv169c3xW7atAmPPPII2rVrh1deeQVubm44e/ZsmclPVbp48SK6du0KT09PzJs3D7m5uXjrrbdw5MgR7Nu3D1qtttQ2ubm5eOGFF+Dq6lqtdSU7JWqZfv36ifDwcJvus6ioSBQWForffvtNABCrVq2qlnJrusLCQuHl5SW6du0qjEajaf13330nAIglS5bcc/urV6+Ko0ePllo/evRoAUCcPn3atM5gMIhWrVqJTp06ifz8/Hvud+3atQKAWL9+veQzqrxVq1aJit4e+fn5wtfXV/Tr189s/fDhw4Wrq6u4ceNGqW1u3bolGjRoIObOnSsAiPj4eJvWu6qEh4eLkSNHmv5et26dACC2b9+uWJ2U8OyzzwoHBwdx4sQJ07q8vDwRGhoq2rdvX+H2O3fuFAaDodQ6AOKll14yW//GG28IR0dHsXfv3nvuMysrSwQGBoqBAweW2rctlffZeKcJEyYIZ2dn8eeff5rWbd68WQAQK1euLHOb6dOni6ZNm5reN0TWqFFJTXZ2tpg8ebIIDw8XWq1W+Pv7i9jYWJGcnCyEEOLBBx8UAMyWkkSjsLBQvPLKK6J9+/bCw8NDuLi4iPvvv19s27bNrIyUlBQBQLz55pti0aJFIjIyUqjVarFo0aJS+y55E9+r3JL93flmHzlypHB1dRV//vmn6Nevn3B1dRUhISHi3//+txBCiMOHD4vu3bsLFxcXERYWJtasWVPqWNy8eVNMnjxZ1K9fX2i1WtGwYUMxf/78Kv3QultycrIAIJYuXVrqMTc3N9G5c2eL9vvtt98KAOLbb781rfvxxx8FAPHDDz8IIW5/URQXF5e5fadOnUTHjh2FELeTodzc3HLLsvQ4Viap+f777wUA8f3335ut3717twAgPvnkk1LbzJkzR4SFhYn8/HybJDV//PGHGDx4sPD29hY6nU5ER0eLb775pszn8uuvv4qpU6cKPz8/4eLiIgYMGCCuXLliFms0GsWrr74q6tWrJ5ydnUW3bt3E0aNHzZKakv3dvZQkOA8++KB48MEHTfvcvn27ACDWrl0rZs+eLUJCQoSbm5sYPHiwyMzMFAUFBWLy5MnC399fuLq6ilGjRomCgoJSz/WTTz4R7du3F05OTsLb21sMHTpUXLhwwarjJ6t169bivvvuK7U+Pj5eABCnTp2yaL8+Pj5i0KBBpr8NBoMICQkRQ4YMEULc/uGVl5dX5rbLly8XAMTx48eFEELk5ube8/Vt6XGsTFITEBAgHnvssVLrmzRpInr27Flq/alTp4RWqxXff/+96XPTUiNGjBA6nc50HErExcUJLy8vcenSpUrva9OmTaJLly7C09NTuLq6iiZNmoiEhASL60bVp0Zdfho/fjy+/PJLTJw4ES1atMD169fx66+/4o8//kD79u3x0ksvISsrCxcvXsSiRYsAwHQNNjs7G++//z6eeOIJjB07Fjk5Ofjggw/Qq1cv7Nu3r9TlpFWrVqGgoADjxo2DTqfDwIEDkZOTg5kzZ2LcuHF44IEHAACdO3dGvXr1yi23PAaDAX369EHXrl2xYMECrFmzBhMnToSrqyteeuklDB8+HIMGDcKKFSswYsQIxMTEICIiAsDtvigPPvggLl26hL///e8ICwvD7t27kZCQgLS0tAr79uTm5qKgoKDC4+3o6AhPT89yHy8sLAQAODs7l3rM2dkZBw8ehNFohFot1zUrPT0dAODn52dat2XLFgCATqdDhw4dkJycDK1Wi4EDB2LZsmXw8fEBcPs879u3D88++yxefPFFvPPOO8jNzUVERATmz5+PIUOGmPZp7XGsyMGDBwEAHTp0MFsfHR0NtVqNgwcP4sknnzStv3DhAubPn48PP/ywzGMq69ixY+jSpQvq1auHGTNmwNXVFV988QUGDBiAr776CgMHDjSLnzRpEry9vTFr1iycP38eixcvxsSJE7F27VpTzMyZM/Haa6+hb9++6Nu3Lw4cOIC4uDjo9XpTTNeuXfGPf/wDS5YswYsvvojmzZsDgOnf8iQmJsLZ2RkzZszAmTNn8M4778DR0RFqtRo3b97E7NmzsWfPHqxevRoRERGYOXOmadvXX38dr7zyCoYMGYJnnnkGV69exTvvvIOuXbvi4MGD8PLyKrfcoqIiZGVlVeqY+vj43PP1XFhYCG9v71LrXVxcAADJyclo3LhxpcoqkZubi9zcXLP3w/Hjx3H58mVERUVh3Lhx+Oijj6DX69G6dWu8/fbb6N69uyl2y5Yt8PDwwKVLlzBgwACcOnUKrq6ueOqpp7Bo0SKzS8TWHMeKXLp0CVeuXCn1fgCAjh074ocffii1fsqUKejevTv69u2LL774wuKyAeDtt9/Gtm3bMHLkSCQlJUGj0WDlypXYtGkTPvnkE4SEhFRqP8eOHcPDDz+MqKgozJ07FzqdDmfOnMF///tfq+pH1UTprOpOnp6eFf5yLe8yUHFxsSgsLDRbd/PmTREYGCiefvpp07qSlhUPD49Sv1ItufxUXksNADFv3jyzujg7OwuVSiU+//xz0/oTJ04IAGLWrFmmda+++qpwdXUt9atvxowZQqPRVPirqqT8ipY7f02X5erVq0KlUokxY8aYrS+pMwBx7dq1e+7jboWFhaJFixYiIiJCFBUVmdY/+uijAoDw9fUVw4cPF19++aV45ZVXhIODg+jcubPp8teBAwdMcYGBgWLZsmVizZo1omPHjkKlUokff/zRtE9rjmNlWmri4+OFRqMp8zF/f3/x+OOPm637v//7P7PWLVjZUtOzZ0/RunVrs1YNo9EoOnfuLBo3bmxaV/JcYmNjzS4jTp06VWg0GpGZmSmEEOLKlStCq9WKfv36mcW9+OKLAkClLz+V11LTqlUrodfrTeufeOIJoVKpRJ8+fcy2j4mJMXuvnT9/Xmg0GvH666+bxR05ckQ4ODiUWn+3kvIrs6SkpNxzX4888ojw8vIS2dnZpeoMQLz11lv33L4sr776qgAgtm7dalq3fv160+u8cePGYtWqVWLVqlWicePGQqvVit9//90UGxUVJVxcXISLi4uYNGmS+Oqrr8SkSZMEALPXoLXHsbzPxhIln58ff/xxqceef/55AcDstbpx40bh4OAgjh07JoQQVrfUCCHEzz//LACI1157TZw7d064ubmJAQMGSO2jpNX+6tWrVtWFlFGjWmq8vLywd+9eXL58udJZdQmNRgONRgPg9p0xmZmZMBqN6NChAw4cOFAqfvDgwfD397dJvcvzzDPPmP7v5eWFpk2b4syZM2atCU2bNoWXlxfOnTtnWrdu3To88MAD8Pb2NrvLKDY2FvPnz8euXbswfPjwcst94YUXzFoIylPWL847+fn5YciQIfjoo4/QvHlzDBw4EJcuXcKkSZPg6OiIoqIi3Lp1q8Jy7jRx4kQcP34c33//PRwc/nr55ebmAgDuu+8+/Oc//wFw+xy5uLggISEBW7duRWxsrCnu+vXr2LNnDzp16gQAePTRRxEREYHXXnsNvXv3BiB3HG/evGl2R1ZJOXff5eXi4mL6VX7r1q0yOz4CgJOTk9mx2b59O7766ivs3btX6niV58aNG9i2bRvmzp2LnJwc5OTkmB7r1asXZs2ahUuXLqFevXqm9ePGjYNKpTL9/cADD2DRokX4888/ERUVhS1btkCv12PSpElmcVOmTMG8efOsrvOIESPMOtR36tQJn332GZ5++mmzuE6dOmHJkiUoLi6Gg4MD1q9fD6PRiCFDhpidj6CgIDRu3Bjbt2/Hiy++WG65bdq0webNmytVxzs76pZlwoQJ+O677zB06FC8/vrrcHV1xbJly7B//34AkH4/7Nq1C3PmzMGQIUPQo0cP0/qS119OTg4OHjyI0NBQAECPHj3QqFEjLFiwwPQ+yc3NRX5+PsaPH2+622nQoEHQ6/VYuXIl5s6di8aNG0sdx/z8fOTn55eqb25urtm2Go3G9DlS8tx1Ol2p7Upai27dugWdTge9Xo+pU6di/PjxaNGihdQxu5e4uDj8/e9/x9y5c/Hll1/CyckJK1eulNpHSWvVN998g9GjR0u3RJPClM6q7rR27Vrh5OQk1Gq1uO+++8SsWbPE2bNnzWLu1WF39erVonXr1sLR0dHs11dERIQppqRlZe7cuaW2t2VLjZOTU6nYBx98UDRr1qzU+vDwcPHwww+b/nZ2dr7nr8mFCxeW+fyrQmZmpqkVpWR58sknxaBBgwQAcfPmzUrva8GCBQKAePXVV0s91q9fPwFAfPTRR2br//zzTwFAzJkzRwjx1zm685yWGD16tHB0dDS1AMkcx/Dw8Er9kr+zRa2yLTVFRUWiVatWYsSIEWYxsKKlZu/evRXW9cCBA0KIv1pq9uzZY7aPkhaMHTt2CCGESExMFABKveeEEMLb29vqlpo7WyjvVa9Zs2aZtQJOmDDhns8zKiqq0sfNFt555x3h6upqKr9Ro0am1/aiRYsqvZ8//vhD+Pj4iLZt25Zq+Sk5vt27dy+1Xffu3c1e/y1bthQAxM6dO83iSjogl7ynZI5jyTmoaLnzM1GmpWb+/PnC29tbXL9+3RRji5YaIYTIyckRQUFBAoD49NNPpbfPz88XXbp0EQCEn5+fGDp0qFi7dm219mcky9WolpohQ4bggQcewIYNG7Bp0ya8+eabeOONN7B+/Xr06dPnntv+5z//wahRozBgwAA8//zzCAgIgEajQWJiIs6ePVsq3hZ9Gu6lpNWosuuFEKb/G41GPPTQQ3jhhRfKjG3SpMk9y87KyqrUL0atVmvqq1IeT09PfPPNN7hw4QLOnz+P8PBwhIeHo3PnzvD396/0NfjVq1dj+vTpGD9+PF5++eVSj5e0zAUGBpqtDwgIAHC7JeVecSWxRUVFyMvLg6enp9RxXLNmjdkxK3n93f0LPzIy0vT/4OBgGAwGXLlyxVRP4PbYItevXzfV9eOPP8bJkyexcuXKUuO65OTk4Pz58wgICDC1AFWG0WgEAPzzn/9Er169yoxp1KiR2d+Vee1VJUvfE0ajESqVCj/++GOZsRX1b9Pr9bhx40al6ujv719ufUpMnDgRo0ePxuHDh6HVatG2bVt88MEHACp+b5ZITU1FXFwcPD098cMPP8Dd3d3s8Ype5yX9uUpijx07VuF7R+Y4jhgxAvfff7/Z4w899BCef/55xMXFmdbd+TkaHBwMAGWOuZOWlgYfHx/odDpkZWXhtddew7PPPovs7GxkZ2cDuN0KJITA+fPn4eLiYvaeknHn0A5HjhzBE088IbW9s7Mzdu3ahe3bt+P777/HTz/9hLVr16JHjx7YtGlTha8PUlaNSmqA22+MZ599Fs8++yyuXLmC9u3b4/XXXzclNXc2i9/pyy+/RGRkJNavX28WM2vWrEqXXd6+K3rM1ho2bIjc3FzExsZatP3kyZNLjQFTlgcffBA7duyo1D7DwsIQFhYGAMjMzERycjIGDx5cqW2/+eYbPPPMMxg0aBCWLl1aZkx0dDTee+89XLp0yWz95cuXAcB0qTAkJARBQUGl4kpinZycTF8QMsexS5cuZn+XjO9xr21LOp/v378fffv2Na3fv38/jEaj6fELFy6gqKioVBnA7YTn448/xoYNGzBgwIAK61miJLlydHS0+HVyt/DwcADA6dOnzZK3q1evmr4YS1T3+0EIgYiIiEonDXfavXu3Wcfae0lJSSk17k5ZXF1dERMTY/p7y5YtcHZ2LvMc3+369euIi4tDYWEhtm7dakoG7tS6dWs4OjqW+zq/89J5dHQ0Nm/ejEuXLqFp06ZmccBf7x2Z4xgZGWn2GijRokWLcl9v9erVg7+/v+lS3J3uvFnj5s2byM3NxYIFC7BgwYJSsREREejfvz++/vrre9axLHl5eRg9ejRatGiBzp07Y8GCBRg4cCDuu+8+qf2o1Wr07NnTND7XvHnz8NJLL2H79u02e79R1agxSY3BYEBubq7Z3TgBAQEICQkx3YUD3P4wKetOhpLsWQhh+sDdu3cvkpKSTF/GFSkZ/CkzM7PMxyp7B4W1hgwZgtmzZ+Pnn38u9Ss8MzMTbm5uZv1R7marPjXlSUhIQHFxMaZOnWq2/sSJE3BxcTE73rt27cLjjz+Orl27Ys2aNeVen+7fvz8mT56MVatWYdSoUaa4999/H8DtX4klhg4dirfffhubN282rb927Rq++eYb9OjRw7SttcexIj169ICPjw+WL19ultQsX74cLi4u6NevHwDg8ccfL3X3HQAMHDgQffv2xdixY019gyorICAA3bp1w8qVKzFp0qRSX4xXr16V7jMWGxsLR0dHvPPOO4iLizO9j8q6S+xe7xVbGzRoEBISEjBnzhz85z//MUuohBC4ceMGfH19y93eln1qyrJ7926sX78eEyZMMPv8SktLQ1ZWFho2bGjqS5SXl4e+ffvi0qVL2L59e7l3Srm7u6Nv377YuHEjTpw4gWbNmgG4PVrv7t278fe//90UO2TIEMyfPx8ffPCBWb+c999/Hw4ODqYpK6w9jpUxePBgfPTRR0hNTTX1A9q6dStOnTpl+rwICAjAhg0bSm27ZMkSJCUl4bPPPisz0auM6dOn48KFC9izZw+aNm2KrVu3YuTIkTh48GCZfX3KcuPGjVIt2CXv3zu/i8r6vLtw4QLy8/NN5wu4/dl07do1hIWFSbXGkoUUu/B1l5s3bwpXV1cxcuRIsXDhQvHuu++KIUOGCADiX//6lymu5Nr11KlTxaeffmoa6+TDDz8UAMSjjz4qVq5cKWbMmCG8vLxEy5Ytza773jlOzd30er3w8vISTZs2Fe+//7747LPPxLlz5+5Z7r3Gqbnbgw8+KFq2bFlqfXh4uNkAbnl5eaJ9+/bCwcFBPPPMM2L58uXirbfeMu23OnvlJyYmiuHDh4slS5aIZcuWibi4ONPdBXcDzO+oOn/+vPD09BTOzs5i6dKl4pNPPjFb7ryDQwhhGpDuoYceEkuXLhXjxo0TKpVKPPHEE2Zx6enpIjg4WLi7u4tZs2aJhQsXiiZNmghnZ2dx6NAhU5w1x7Eydz8JIcTSpUsFAPF///d/4r333hMjRowQACq8k6TkeJXVp6ZkXKSKHDt2THh7ewtfX18xY8YM8e6774pXX31V9O3b16x/RMlz+e2338y2L+nrcme/mISEBAFA9O3bV/z73/8WY8aMESEhIcLPz8+sT01aWprQaDTib3/7m1i9erX47LPPREZGhqn+ZfWpWbdunVn55dWrpD/HneenpL9P586dxYIFC8Ty5cvFCy+8IBo3blzme7mqnD9/XnTs2FG89tpr4v333xdTp04Vzs7Ool27dqX6xZTchXjnHVX9+/cXAMTTTz9d6v2wYcMGs+2PHTsm3NzcRHBwsEhMTBSJiYkiODhY+Pv7i4sXL5rFPv300wKAGDJkiFi6dKl47LHHBIBSY6tYcxzv/pwry4ULF4Svr69o2LChWLJkiZg3b57w9vYudZdeWcr73Cx5nVRU9tatW4VKpRKzZ882rdu1a5dQq9Xi+eefv+e2d5o8ebJo166dePnll8V7770nXn/9dVGvXj1Rv359052CQpT+vBOi7Pduyeu5rg1UqZQak9QUFhaK559/XrRp00a4u7sLV1dX0aZNG7Fs2TKzuNzcXDFs2DDh5eVl1lHNaDSKefPmifDwcKHT6US7du3Exo0bxciRIyud1AghxDfffCNatGghHBwczN5I5ZVbFUmNELc7uyUkJIhGjRoJrVYr/Pz8ROfOncVbb71ldltsVdu4caPo2LGjcHd3Fy4uLuJvf/ub+OKLL8qMvftNXtGttHd2uhXi9jl85513RJMmTYSjo6MIDQ0VL7/8cpnP9+zZs2LgwIHCw8NDODs7ix49eoh9+/aVirP0OFY2qRFCiHfffVc0bdrUNLjfokWLzG6JLk95SU10dLQICgqqVNlnz54VI0aMEEFBQcLR0VHUq1dPPPzww+LLL78s9Vwqk9QYDAYxZ84cERwcXO7geyXee+89ERkZKTQajdl+qiKpEUKIr776Stx///3C1dVVuLq6imbNmon4+Hhx8uTJSh0rW7hx44bo37+/CAoKElqtVkRERIjp06eXSmiEKDupuVeH9LJuREhOThaxsbHC1dVVuLu7i/79+5c5wJ9erxezZ88W4eHhwtHRUTRq1KjcTsuWHsfKJBZCCHH06FERFxcnXFxchJeXlxg+fLhIT0+vcLvyPjffeecdAUD89NNP5W6bnZ0twsPDRfv27c2GihDi9tAFarVaJCUlVVgHIW4nR/379xchISFCq9WKkJAQ8cQTT5Q67kxqaiaVENXUS5CIKpSTkwMfHx8sXrwY8fHxSleHSHFDhgzB+fPnsW/fPqWrQrVAjelTQ0S3+yDVq1cPY8eOVboqRIoTQmDHjh2mMXmIKsKWGiIiqhNKpmgpj7Oz8z2njqGaj0kNERHVCRUNRTBy5EisXr26eipDVYKXn4iIqE6o6NZ+2el5qOZhSw0RERHZBc7URURERHbB7i8/GY1GXL58Ge7u7tU6tDsREdU+Qgjk5OQgJCSkSmfoLigogF6vt3o/Wq3WNAs61YGk5vLly6bhuomIiCojNTUV9evXr5J9FxQUwNfZDfkwWL2voKAgpKSkMLH5H7tPakomNzzy+gS4O1Vu7g+Xpi2ky8k99rtUfOaZ0hPVVaQ4r0B6Gxl/bEqR3sZJI/dLJrBt6VmHK+IaIDdfirOfl1S8e7j8fD8uTVtKxat08nO+qLSVe72WUDu7SsULB/kPQf3R/0rFn1u/TboMZz8PqfjgbnLzZqnvHyIVDwAO6Sek4o3ZlZsV/E6q4IZS8V9elTtOuUXFUvEA0DrAveKgO6w7dFm6DLVargW9S6RPxUF3SM26JRVfkJeLV//vgVIzp9uSXq9HPgwYjnrQWtELRA8j1qRfgl6vZ1LzP3af1JRccnJ30sHDuZJJjasFX0CVTJhKGLSO0mUUFVmf1d+Li1ojvY2TZPOsm6P8S85V8li56LRS8ZV9XZiVIfkaUTlZktTIfUhJJzWOzlLxAKB3kauTm6P869xFJ7eNh2Sd1B5yyQAAOOTKHVujQf4HiMrdTSre+Zbcl65BL5/UuEp+sWtd5J4DAGgkkxoXN7k6ORVb9jVXHd0VnKGGVmV5UqPhbT6l2H1SQ0REVBNpVCporEieNFDdnjmMTJjUEBERKUCtAjRWNAipASY1d+Et3URERGQX2FJDRESkAJtcfiIzTGqIiIgUoLHy8pP8rR32j5efiIiIyC6wpYaIiEgBvPxke0xqiIiIFMDLT7bHy09ERERkF+pMS417dGe4u1VuVFdDTqb0/o2So3Xeup4rXYZBLzeicPbFHKl42SkPACAgyl8uvk2YdBkaJ7kRgr0ay5XhEBIhFQ/IjxAsO9ovID+isCHrulz89XSpeAC4uvewVHzxLflRbB1d5EY6VrvIjTBryUjK0MmdP3WQ/KjF+T6RUvG5FzOk4rNuFUnFA8DXR+RfI7JkRxS+kic3CeTFG3LTJBTmV+10NHfi5SfbU7SlZvny5YiKioKHhwc8PDwQExODH3/80fR4QUEB4uPj4evrCzc3NwwePBgZGXJvZCIioppIhdtfwpYuTGlKUzSpqV+/PubPn4/k5GTs378fPXr0QP/+/XHs2DEAwNSpU/Hdd99h3bp12LlzJy5fvoxBgwYpWWUiIiKqoRS9/PTII4+Y/f36669j+fLl2LNnD+rXr48PPvgAn376KXr06AEAWLVqFZo3b449e/bgb3/7mxJVJiIisglefrK9GtOnxmAwYN26dcjLy0NMTAySk5NRVFSE2NhYU0yzZs0QFhaGpKSkcpOawsJCFBYWmv7Ozs6u8roTERHJ4t1Ptqf43U9HjhyBm5sbdDodxo8fjw0bNqBFixZIT0+HVquFl5eXWXxgYCDS08vvvJaYmAhPT0/TEhoaWsXPgIiISN7tpEZlxaL0M6h5FE9qmjZtikOHDmHv3r2YMGECRo4ciePHj1u8v4SEBGRlZZmW1NRUG9aWiIiIairFLz9ptVo0atQIABAdHY3ffvsNb7/9NoYOHQq9Xo/MzEyz1pqMjAwEBQWVuz+dTgedTlfV1SYiIrIKLz/ZnuItNXczGo0oLCxEdHQ0HB0dsXXrVtNjJ0+exIULFxATE6NgDYmIiKxn3aUn6zoZ2ytFW2oSEhLQp08fhIWFIScnB59++il27NiBn3/+GZ6enhgzZgymTZsGHx8feHh4YNKkSYiJieGdT0RERFSKoknNlStXMGLECKSlpcHT0xNRUVH4+eef8dBDDwEAFi1aBLVajcGDB6OwsBC9evXCsmXLlKwyERGRTaitvPxU4y611ACKJjUffPDBPR93cnLC0qVLsXTpUusLc3AAHBwrFSr08sNk56XfkIrPPJ8lXUbqmZvS28ioH+klv839TaXifTq0ky5D7So55Hwlz7Np/5LTEQCAMV9uCgqjdAnyH1g3d26tOOgO+px8yRKA/Ktyr8H6D7aULsMtQnKai0ZyrylDgfx7T/Y1VewtPx3IzQK5aVA8dXIf34808ZOKB4D3frsoFZ8vOZULAOiL5d4dxy7JDdNxXwNvqfhbudWXKnCcGttjokdERFRHXLp0CU8++SR8fX3h7OyM1q1bY//+/abHhRCYOXMmgoOD4ezsjNjYWJw+fVrBGsthUkNERKSAkrufrFlk3Lx5E126dIGjoyN+/PFHHD9+HP/617/g7f1Xa9aCBQuwZMkSrFixAnv37oWrqyt69eqFgoLqm+jTGorf0k1ERFQXVfct3W+88QZCQ0OxatUq07qIiAjT/4UQWLx4MV5++WX0798fAPDxxx8jMDAQX3/9NR5//HHLK1tN2FJDRERUB3z77bfo0KEDHnvsMQQEBKBdu3Z47733TI+npKQgPT3dbHoiT09PdOrUCUlJSUpUWRqTGiIiIgXYapya7Oxss+XO+Q/vdO7cOSxfvhyNGzfGzz//jAkTJuAf//gHPvroIwAwTUEUGBhotl1F0xPVJExqiIiIFKCBlX1q/ref0NBQszkPExMTyyzPaDSiffv2mDdvHtq1a4dx48Zh7NixWLFiRbU956rGPjVEREQKUFt5S7f6f9umpqbCw+OvoS/KmyooODgYLVq0MFvXvHlzfPXVVwBgmoIoIyMDwcHBppiMjAy0bdvW4npWJ7bUEBER1WIeHh5mS3lJTZcuXXDy5EmzdadOnUJ4eDiA252Gg4KCzKYnys7Oxt69e2vN9ERsqSEiIlKA1Xc/SW47depUdO7cGfPmzcOQIUOwb98+vPvuu3j33XcBACqVClOmTMFrr72Gxo0bIyIiAq+88gpCQkIwYMAAyytajZjUEBERKcDqEYUlt73vvvuwYcMGJCQkYO7cuYiIiMDixYsxfPhwU8wLL7yAvLw8jBs3DpmZmbj//vvx008/wclJfuR1JTCpISIiqiMefvhhPPzww+U+rlKpMHfuXMydO7caa2U7dSepKS4GiosqFWq4nia9e0NB2bfQlSfvivy8O6dy9VLxuZJzqtw6LTd/FQBEFMjVqfi6/G2B+UePScU7ujpLxTvVC5GKB4DiTLljpa0fKV2GMeu6VHxOaoZUvD5b/jXo0zxcKt7r/u7SZahCGsvFF8jNw5XvKDmXGAA3jdx8Uapiuc8DAEi5KdfFMci97H4T5fnsiPx7L8RL7r10XfIzyhIuWtkh52qu6r78VBfUnaSGiIioBqnuy091Ae9+IiIiIrvAlhoiIiIFqFUq01gzlm5P5pjUEBERKUClUUGltjwxUTGpKYWXn4iIiMgusKWGiIhIAWqNCmorWmp4+ak0JjVERERK0KihUltxwUQlbFcXO8GkhoiISAEqtQoqKwabUYEtNXdjnxoiIiKyC2ypISIiUoBao4LaipYaNVtqSqkzSY3a3RNqN9dKxWo8faX3r/Nyl4oXBrkpDAD5ZjXZaRJyi+Wvz6od5V5C1w6flS6jOO+WVLxKI3eknK7clIoHAJcAb6l4rdEgXYaxQG4aA2d/uTo5+XhKxQOAW2igVLzaXa5OAGDUVu59aoqX3L+zSv5cpDvVk4q/kS9fhouj3DNJy5WbiiErv3LTxNzJw8lRKt7LRS4eADIl61Uo+bl2Il1uGo3C/FypeGuo1Nb1qVEJ9qm5Gy8/ERERkV2oMy01RERENQkvP9kekxoiIiIFqDS8+8nWePmJiIiI7AJbaoiIiBRwu6XGio7C0t3k7R+TGiIiIgWwT43t8fITERER2QW21BARESlApVJBZcWEliojW2ruxqSGiIhIAWqNGmor+tSoBS+23I1JDRERkQKsvqVbsKXmbkzziIiIyC7UnZYatQOgqdy8JBp/uXleAMA54IxUfOOHm0qXce3zI1LxQU5yp9dHq5GKBwB9jtz8RJYouJknFe/fJlIqXnbOJEB+fimNd4B0GY7uXlLxAdFxUvFGJ7n5ygBAdWaf3AYWlGFwlZt7TSPkbmvVZF2WigcAf88Qqfhrt6p+Th6d5GvQTfLzAADOX5N77zX0l5u3CwCy3Yql4s9dkatTToHc/vWS8dZgS43t1Z2khoiIqAZhnxrb4xEhIiIiu8CWGiIiIiVYefkJvPxUCpMaIiIiBahVKqitGKdGrWJSczdefiIiIiK7wJYaIiIiBag0ausmtDSyXeJuTGqIiIgUYPWElpwmoRSmeURERGQX2FJDRESkAKsH32NLTSlMaoiIiBTAPjW2V2eSmqKAxijy8KhUrNrNX3r/rmnnpeKdfD2ly2h1MVMqPictVyreO9JLKh4A8tJuSMUXWzAEuZO3s1y85LHVhTeSigcAjW+QVLza3Ue6DMN1yeH8s65LhVvyG08dEC4VXxTQRLoMIXmbqspQJF2GNMmpGAqK5eIBIMC1ctO4lHCU/IWvseDWYdlpD7yd5Z6DJdvcyNVLxWfmV8Prw0JqDazsU2PDytgJpnlERERkFxRNahITE3HffffB3d0dAQEBGDBgAE6ePGkW061bN6hUKrNl/PjxCtWYiIjINlRqldULmVM0qdm5cyfi4+OxZ88ebN68GUVFRYiLi0NenvksrGPHjkVaWpppWbBggUI1JiIisg21Wm2a1NKiRc2LLXdTtE/NTz/9ZPb36tWrERAQgOTkZHTt2tW03sXFBUFBcn0YiIiIqG6pUWleVlYWAMDHx7xT5Zo1a+Dn54dWrVohISEB+fn55e6jsLAQ2dnZZgsREVFNU3JLtzULmasxdz8ZjUZMmTIFXbp0QatWrUzrhw0bhvDwcISEhODw4cOYPn06Tp48ifXr15e5n8TERMyZM6e6qk1ERGQRq2/ptmJbe1Vjjkh8fDyOHj2Kzz//3Gz9uHHj0KtXL7Ru3RrDhw/Hxx9/jA0bNuDs2bNl7ichIQFZWVmmJTU1tTqqT0REVGvMnz8fKpUKU6ZMMa0rKChAfHw8fH194ebmhsGDByMjI0O5SlqgRiQ1EydOxMaNG7F9+3bUr1//nrGdOnUCAJw5c6bMx3U6HTw8PMwWIiKimkalVlu9WOK3337DypUrERUVZbZ+6tSp+O6777Bu3Trs3LkTly9fxqBBg2zxVKuNokmNEAITJ07Ehg0bsG3bNkRERFS4zaFDhwAAwcHBVVw7IiKiqmPVnU//W2Tl5uZi+PDheO+99+Dt7W1an5WVhQ8++AALFy5Ejx49EB0djVWrVmH37t3Ys2ePLZ92lVI0qYmPj8d//vMffPrpp3B3d0d6ejrS09Nx69YtAMDZs2fx6quvIjk5GefPn8e3336LESNGoGvXrqUyTCIiIrq3+Ph49OvXD7GxsWbrk5OTUVRUZLa+WbNmCAsLQ1JSUnVX02KKdhRevnw5gNsD7N1p1apVGDVqFLRaLbZs2YLFixcjLy8PoaGhGDx4MF5++WUFaktERGRDVnYUxv+2vfsuX51OB51OVyr8888/x4EDB/Dbb7+Veiw9PR1arRZeXl5m6wMDA5Genm55HauZokmNEOKej4eGhmLnzp22KcxouL1UgsFDfkwcdbcnpeIdbmVJl1H/yk2p+MxTcp2kL+6R71RdcLNAKt7Zz0W6jPCIAKl4XWjFlzHv5BAUJhUPACqtk1S80Vl+ri9Vfblt9F6hUvG3LJifyNVB7hbS/OJ7v8fLotXIxUvP1aaWLADALSG3TbCbfBnBIlNuAxe5j+8b9eRfg+k5hVLxBvnTjbp8V7JKbeXdT//rUxMaav7enzVrFmbPnm22LjU1FZMnT8bmzZvh5CT3+VWb1JhbuomIiOoSazr7lmwP3E5Y7rwppqxWmuTkZFy5cgXt27c3rTMYDNi1axf+/e9/4+eff4Zer0dmZqZZa01GRkatGvyWSQ0REVEtVpk7fXv27IkjR46YrRs9ejSaNWuG6dOnIzQ0FI6Ojti6dSsGDx4MADh58iQuXLiAmJiYKqu7rTGpISIiUsDtwffkL1X+tX3lulQAgLu7u9nAtgDg6uoKX19f0/oxY8Zg2rRp8PHxgYeHByZNmoSYmBj87W9/s7iO1Y1JDRERkQJq2ojCixYtglqtxuDBg1FYWIhevXph2bJlNi2jqjGpISIiqoN27Nhh9reTkxOWLl2KpUuXKlMhG2BSQ0REpAC1Wg21FR2FrdnWXjGpISIiUkBNu/xkD3hEiIiIyC6wpYaIiEgBbKmxPSY1REREClCprBx8T8Wk5m48IkRERGQX6kxLjfr0HqjdKjfvkKpeU+n9n9EES8WH7fhcugyvzg9Kxasd/ysV//OnRyoOukuu5PxBbS1oLhWGyg8wBQCOIQ3k9u8hOXcQABiKpcJVNy/Jl+HqLRWuFnLHqUDuKQAADEa5yX28RJ50GX/ecpaKd3aQ+xjzcZb/2NNJzk/kpZKf0OgWfKXiiyXPRaiH3OsDAJwc5N6vZ2/cki5jb8oN6W3sBS8/2V6dSWqIiIhqEiY1tsekhoiISAFqjRpqKxITa7a1VzwiREREZBfYUkNERKQAlVpl3d1Pavm+W/aOSQ0REZEC2KfG9nhEiIiIyC6wpYaIiEgBbKmxPSY1RERECuCIwrbHI0JERER2gS01REREClBpNFBrNFZtT+bqTFJjyLwGQ1Hlhl/X+oZI7z/PMVAqXqVzki4Dks2UV5JPSsVrLBjaXSt5S6HOQytdRlDntlLxhpxMuQJuXpWLB6AJbSa3gbOHdBkGd7npG7KL5PYvO8w+APx4+rpU/IimrtJlRBjSpeKNJ5Ol4kWRXioeABwatpGKV3uHSpdR5CA3PYRGI/fe89TJN8xnF8p9aRqF/GuqbainVPzxtBypeL3kVC7ViX1qbI9HhIiIiOxCnWmpISIiqknYUmN7TGqIiIgUoFJbefeTFdvaKyY1RERECmBLje3xiBAREZFdYEsNERGRAlRqlXUtNZzQshQmNURERApgnxrb4xEhIiIiu8CWGiIiIgWo1Bqo1FaMKGzFtvaKSQ0REZES1JrbizXbkxlefiIiIiK7UGdaahzrRcDRrXLz0Bjc/KT3f/xirlR8VL2G0mWo6jWVihcGuXlY2rQJkIoHAJ/GPlLxYb06S5fhGNZEKr7wzGGpeF2jKKl4ACj2lzt/Vwvl71I4lJEnFd8zQu5X29wtp6TiAWBuXCOpeAH5eZYKd6yVii+4niUVr/N2l4oHAMf6cudbLzmPEwA4Z16Qii/wCpPbv6P8b1hPJ7nXVKin/Jx2J6/Jvc51DnLPoybP/QS1WnpOv1Lbk5k6k9QQERHVJCqNxqqZtjlLd2lM84iIiMgusKWGiIhICewobHNMaoiIiJSgVluZ1PBiy92Y1BARESmAIwrbHo8IERER2QW21BARESlBZWWfGhX71NyNSQ0REZES2FHY5nj5iYiIiOwCW2qIiIgUwI7Ctldnkpr8sA5w8PCoVOzV/GLp/T/UUO7Fpc4Mly7DqJY7XaGx0VLxET7+UvEAoPENkoovviA/NH/W3l+k4p18PaXi1e5eUvEAoEk7JhUfeD1duoy+3nLnQ5wpkIpfGHRdKh4ANKk3pOJFkfw0CZmpGdLbyHCNjJTexuAdKhW/OzVHuozm/vWk4p2K5Ib/l51ewBKyUx4AwNWcwiqoSS3By082xzSPiIiI7IKiSU1iYiLuu+8+uLu7IyAgAAMGDMDJkyfNYgoKChAfHw9fX1+4ublh8ODByMio2l9yREREVa5k8D2LF7mv8LrwnatoUrNz507Ex8djz5492Lx5M4qKihAXF4e8vL+aMKdOnYrvvvsO69atw86dO3H58mUMGjRIwVoTERFZr2RCS2sWGXXhO1fRPjU//fST2d+rV69GQEAAkpOT0bVrV2RlZeGDDz7Ap59+ih49egAAVq1ahebNm2PPnj3429/+pkS1iYiIap268J1bo/rUZGVlAQB8fHwAAMnJySgqKkJsbKwpplmzZggLC0NSUlKZ+ygsLER2drbZQkREVOOo1dYvQKnvvMLCynW+tsV3bk1TY5Iao9GIKVOmoEuXLmjVqhUAID09HVqtFl5eXmaxgYGBSE8v+26SxMREeHp6mpbQULm7FoiIiKqFVf1p/rpzKjQ01Ox7LzExscKibfWdW9PUmFu64+PjcfToUfz6669W7SchIQHTpk0z/Z2dnc3EhoiIahyVWgOVFbdll2ybmpoKjzuGLNHpdBVua6vv3JqmRiQ1EydOxMaNG7Fr1y7Ur1/ftD4oKAh6vR6ZmZlmmWNGRgaCgsoeH0Wn01XqhBIREdkDDw8Ps6SmIrb8zq1pFL38JITAxIkTsWHDBmzbtg0RERFmj0dHR8PR0RFbt241rTt58iQuXLiAmJiY6q4uERGR7ais7E+jkvsKrwvfuYq21MTHx+PTTz/FN998A3d3d9M1O09PTzg7O8PT0xNjxozBtGnT4OPjAw8PD0yaNAkxMTG1ohc2ERFReWx1+amy6sJ3rqJJzfLlywEA3bp1M1u/atUqjBo1CgCwaNEiqNVqDB48GIWFhejVqxeWLVtWzTUlIiKq3erCd65KCCGUrkRVys7OhqenJ/afSYWbe+WuOZ6/KTeHDgB0CXWXir9+yyBdRtDv66XiDZLzDRn18s/bUCA3t8+V/Seky1Br5XLvwI4tpOKdOz8sFQ8AwlGy31bGeekyDDevSMZflYpX6Zyk4gFA5eAoFS+McvMTAUDh5UtS8W7RXaTiRWhLqXgAMLr6SsWrj2+XLgOSx0oV2lwqPt1Jbm4pALhVLPf14Oygki5D1vzt56p0//r8XLw3oguysrKk+qnIKPleur7rS3i4uVq+n9w8+Hb9vyqta21TIzoKExER1Tl3jDVj8fZkhkeEiIiI7AJbaoiIiBRgyfxNd29P5pjUEBERKeGOUYEt3p7MMKkhIiJSApMam2OfGiIiIrILbKkhIiJSgEqthsqKO5is2dZeMakhIiJSgsrKy08qXn66G9M8IiIisgtsqSEiIlKCSiU9KWWp7clMnUlqbhUZoSmq3DDk3x2Vm17Akm3+1StcuoznbraVin+j+FupeF3sU1LxAJA6/yWp+KvH5Yb+BwCP+nJTUAiD/ND8ssSl01VehiHrulR8TkqqVLzOS+64AoDWU24bx0j5KQl0LTpJxevP/C4Vr5KcfgIARJfHpeIzf5WfJsHBSW7qDTfJ/ftFhUluAdwslHsvaSz4jr1RID9ljN1Qyc+0XWp7MsMjQkRERHahzrTUEBER1SRCpYaworXFmm3tFZMaIiIiJfDyk83xiBAREZFdYEsNERGRElQq6+5g4t1PpTCpISIiUoJafXuxZnsyw6SGiIhIAewobHtMaoiIiKhaHD58uNKxUVFR0vtnUkNERKSEOnj3U9u2baFSqSCEgKqCPkEGg/zAjLXviBAREdmDkqTGmqWWSUlJwblz55CSkoKvvvoKERERWLZsGQ4ePIiDBw9i2bJlaNiwIb766iuL9s+WGiIiIqoW4eF/TRH02GOPYcmSJejbt69pXVRUFEJDQ/HKK69gwIAB0vuvM0nNx79dhNalcrOlaNTyt8kZjEIqXq/WSpfRur6nVLyjvpFUvNA6S8UDgD4nXyreoK/6eV6yU9Kk4p16eUuXodJKzh9kwV0KKkf514gMS+bIEka586dSa6TLMDrJzS/l0LqrXAFC/nkbDHqp+Lz0G9JlOPt6SMUb87Kl4tUFWVLxAODtJPeZ43DzgnQZBU71pLeRIft5rrbg899idfDy052OHDmCiIiIUusjIiJw/Phxi/ZZu48IERFRLSVUKtMdUJYttXucmubNmyMxMRF6/V8/GvR6PRITE9G8eXOL9llnWmqIiIio5lixYgUeeeQR1K9f33Sn0+HDh6FSqfDdd99ZtE8mNUREREqo45efOnbsiHPnzmHNmjU4ceIEAGDo0KEYNmwYXF1dLdonkxoiIiIlcJoEuLq6Yty4cTbbH5MaIiIiqhbffvst+vTpA0dHR3z77bf3jH300Uel98+khoiISAl18PLTgAEDkJ6ejoCAgHvesq1SqSwafI9JDRERkQLq4txPRqOxzP/bSu07IkRERPZApf5rpm5LllqY1JQoKipCz549cfr0aZvut/YeESIiIqqVHB0dpSa3rCwmNUREREqog3M/3enJJ5/EBx98YNN91pk+NQ0DXeHkWrlpEk6n51ZxbQCtoVB6myMX5YY5V0d6ScUX6+SGpwcArbuLVLy3ZJ0AwDMiSCreP7anVLyQHJYfANTecnVSGYuly9BGtpSK9/aXG27ekikMDDclp4ewwC2vMKn4XL3cdXkftfx776ZB7qPSNchHugwnX7kpCVTOcuN4ONyQn8IgJ6i1VHyqQ4h0GemZ8udDhuwUNkbJeKvUwY7CdyouLsaHH36ILVu2IDo6utTYNAsXLpTeZ51JaoiIiKjmOHr0KNq3bw8AOHXqlNljKgvH4GFSQ0REpIQ63lKzfft2m++TSQ0REZECSia0tGZ7MsekhoiIiKrdwIEDy7zMpFKp4OTkhEaNGmHYsGFo2rRppfdZu9uuiIiIaqs6fveTp6cntm3bhgMHDkClUkGlUuHgwYPYtm0biouLsXbtWrRp0wb//e9/K71PttQQEREpoY5PaBkUFIRhw4bh3//+N9Tq2wma0WjE5MmT4e7ujs8//xzjx4/H9OnT8euvv1Zqn7U7zSMiIqJa6YMPPsCUKVNMCQ0AqNVqTJo0Ce+++y5UKhUmTpyIo0ePVnqfTGqIiIiUoNDlp6VLl6JBgwZwcnJCp06dsG/fPhs/scopLi7GiRMnSq0/ceKEaTJLJycnqdu7pY/IyJEjsWvXLtnNiIiI6A4lE1pas8hau3Ytpk2bhlmzZuHAgQNo06YNevXqhStXqn5gzbs99dRTGDNmDBYtWoRff/0Vv/76KxYtWoQxY8ZgxIgRAICdO3eiZcvKD0Qq3acmKysLsbGxCA8Px+jRozFy5EjUqyc3kikREVGdp8A4NQsXLsTYsWMxevRoAMCKFSvw/fff48MPP8SMGTMsr4sFFi1ahMDAQCxYsAAZGRkAgMDAQEydOhXTp08HAMTFxaF3796V3qf0Efn6669x6dIlTJgwAWvXrkWDBg3Qp08ffPnllygqKpLdHREREVkhOzvbbCksLHvqCb1ej+TkZMTGxprWqdVqxMbGIikpqbqqa6LRaPDSSy8hLS0NmZmZyMzMRFpaGl588UVoNLencQkLC8Off/5Z7nO6m0V3P/n7+2PatGmYNm0aDhw4gFWrVuGpp56Cm5sbnnzySTz77LNo3LixJbuuMvU9nOHi5lyp2M6h3tL7P5CWLRX/3Cb5eVgmdG4gFV/s2VAqXiXk5tABgKDObaXiNd7+0mXAQStXhm+wVLzIuy4VDwDi+iWp+KuNukmX4bTldan4izt/l4r3bdVAKh4AfGI6S8UbI6Oly8gulHsdBmUckIo3BMi9LwAgIPOcVPwNjfwv6GuHz0rFu165KRXv/UB3qXgAcPaUm8vps9/zpMt4/oFwqfgreX5S8dtPXZOKr063B9+z/A6mkm1DQ0PN1s+aNQuzZ88uFX/t2jUYDAYEBgaarQ8MDCyzb0t18vDwKPexPn364NChQ4iMjKxwP1Z1FE5LS8PmzZuxefNmaDQa9O3bF0eOHEGLFi2waNGiCrfftWsXHnnkEYSEhEClUuHrr782e3zUqFGme9dLFplmKCIioppKCOsXAEhNTUVWVpZpSUhIUPaJ2ZgQlZ9kVLqlpqioCN9++y1WrVqFTZs2ISoqClOmTMGwYcNMmdaGDRvw9NNPY+rUqffcV15eHtq0aYOnn34agwYNKjOmd+/eWLVqlelvnU4nW2UiIiK75eHhcc+WjhJ+fn7QaDSm/islMjIyEBQUVFXVq1bSSU1wcDCMRiOeeOIJ7Nu3D23bti0V0717d3h5eVW4rz59+qBPnz73jNHpdHZzsImIiEoYhYBRohWirO1laLVaREdHY+vWrRgwYMDtfRiN2Lp1KyZOnGhxPWoS6aRm0aJFeOyxx+Dk5FRujJeXF1JSUqyqWIkdO3YgICAA3t7e6NGjB1577TX4+vraZN9ERERKEf9brNle1rRp0zBy5Eh06NABHTt2xOLFi5GXl2e6G6q2k05qnnrqqaqoR5l69+6NQYMGISIiAmfPnsWLL76IPn36ICkpydQz+m6FhYVmvaSzs+U68BIREdmroUOH4urVq5g5cybS09PRtm1b/PTTT6U6D9ckMoPv1ei5nx5//HHT/1u3bo2oqCg0bNgQO3bsQM+ePcvcJjExEXPmzKmuKhIREVnEKG4v1mxviYkTJ9aqy00yHYVr1TQJkZGR8PPzw5kzZ8qNSUhIMOsFnpqaWo01JCIiqhwhhNVLbTZ37lxs27at1Pq8vDzMnTvX9HdOTk6lbucGallSc/HiRVy/fh3BweWPQ6LT6Uw9wSvbI5yIiIiq1+zZs9GnTx8sXLjQbH1ubq7FV1wUvfyUm5tr1uqSkpKCQ4cOwcfHBz4+PpgzZw4GDx6MoKAgnD17Fi+88AIaNWqEXr16KVhrIiIi6yl1+akm+fjjjxEfH48jR45g5cqV0GrlBlu9m6ItNfv370e7du3Qrl07ALd7Zbdr1w4zZ86ERqPB4cOH8eijj6JJkyYYM2YMoqOj8csvv3CsGiIisgvCisUedO/eHXv37sXevXvRrVs3qyfWVLSlplu3bve8Jvjzzz/brKydp69B61JQqdhZDzWS3n9DyeH/+zWWvy397M3K1b9EQy9XqXhN5kWpeAAwOjpKxeefPildhs7HUype06SDVLzKoJeKB4DCc8ek4l0O75Uu41LSUan4i3suS8XnpOVKxQOAz31yx/bPIrnXIABEXPpFKt5QIDc0v7hYfp+88qTv2ScVr8/Jly4jK0Xuw1yfLfe8XevJvZ4AwCmwgVR8Qlsf6TKMBrk5Ax8Ik/s8qMnTJNT1lpqSu5oaNmyIPXv2YMiQIYiOjsaKFSss3met6lNDRERE9uHORg0PDw/88MMPGDhwoGlgQEvU6Fu6iYiI7JW1dzDV9rufVq1aBU/Pv1re1Go1lixZgnbt2mHXrl0W7ZNJDRERkQKM/1us2b42GzlyZJnrR48ebfEIx7z8RERERIrYunUrHn74YTRs2BANGzbEww8/jC1btli8PyY1REREChDC+qU2W7ZsGXr37g13d3dMnjwZkydPhoeHB/r27YulS5datE9efiIiIlJAXb/7ad68eVi0aJHZlA3/+Mc/0KVLF8ybNw/x8fHS+2RLDREREVW7zMxM9O7du9T6uLg4ZGVlWbRPJjVEREQKqOtzPz366KPYsGFDqfXffPMNHn74YYv2yctPRERECqiLdz8tWbLE9P8WLVrg9ddfx44dOxATEwMA2LNnD/773//iueees2j/TGqIiIioWixatMjsb29vbxw/fhzHjx83rfPy8sKHH36Il19+WXr/TGqIiIgUIGDdHUy18eJTSkpKle6/ziQ1HcK94ezmXqlYrUYlvX8nwy2peFetRroMnb+LVPz0n+TmuDFY0JV+0f0DpOJdjuyQLkPTrKNUfLFXfal4x3N7pOIBAEaDVPjVQ6eki8jLkJvbp/hWsVS8xoLXoLFAbk6jhpmH5ctQy3X1E4Vyc6Jd+6/8+c4+nyYVr3aU/2jVusrNTuwaLDd/nGNYE6l4AFAVyx1bTXa6dBmyMorl5n6qyYxCwGhFVmPNtkqZNm1apeJUKhX+9a9/Se+/ziQ1RERENYm1s23XvpQGOHjwoNnfBw4cQHFxMZo2bQoAOHXqFDQaDaKjoy3aP5MaIiIiqhbbt283/X/hwoVwd3fHRx99BG9vbwDAzZs3MXr0aDzwwAMW7Z+3dBMRESmgZPA9a5ba7F//+hcSExNNCQ1wu+Pwa6+9ZtGlJ4AtNURERMqwdqqDWp7UZGdn4+rVq6XWX716FTk5ORbtky01REREVO0GDhyI0aNHY/369bh48SIuXryIr776CmPGjMGgQYMs2idbaoiIiBRghIDRiuYWa7atCVasWIF//vOfGDZsGIqKigAADg4OGDNmDN58802L9smkhoiISAHWzrRdC+/oNuPi4oJly5bhzTffxNmzZwEADRs2hKurq8X7ZFJDREREinF1dUVUVJRN9sWkhoiISAHW3sFU2+9+qgpMaoiIiBRQ1y8/VQXe/URERER2oc601OQWFcOgr9zcOLeK5Sd0dzbq5TaQmzoIAOCkk5sbJibSRyp+cJDkcwBgdPOTildHdZcuI9clQCreSSU3d5fKSb5TmkPcaKn44q3PSpfh4ic311ebp+XmyPJqEioVDwC65vdJxRenn5cuQ+MpN6eRg1+QVHxhpvz4FyqN3O8/B2e59yoAuNXzl4qXPX+akEZS8QBgdKrcfHlWEXKftyt3/1lFFal+df3up6pQZ5IaIiKimoSXn2yPSQ0REZEC6uIs3VWNfWqIiIjILrClhoiISAEG4+3Fmu3JHJMaIiIiBfDyk+3x8hMRERHZBbbUEBERKcAoBAxsqbEpJjVEREQKuD1NgjVJjQ0rYyd4+YmIiIjsAltqiIiIFMC7n2yvziQ13Rv4wM3do1KxqVny0wV4+0oOtW8oki7DINnWmJFbKBU/9ddcqXgAGNTGTSr+5DXpIjCwudzzdtLITZNg8JAbZh8A/iySO9/eFkxJ4Dv0GeltZKj0t6S3KfYMkYrP9GspXYa7TiMVfyFb7v3qG/GDVDwAOPt7S28jS+MbXKX7v+HbTHqb/CK5b02dg9x7DwASt52Tind3qtqvrUJj9X0t8u4n2+PlJyIiIrILTGqIiIgUYPjf3U/WLFXh/PnzGDNmDCIiIuDs7IyGDRti1qxZ0OvNW0UPHz6MBx54AE5OTggNDcWCBQuqpD4y6szlJyIioprECOvuYKqqLjUnTpyA0WjEypUr0ahRIxw9ehRjx45FXl4e3nrrLQBAdnY24uLiEBsbixUrVuDIkSN4+umn4eXlhXHjxlVRzSrGpIaIiEgBBqOQ7it59/ZVoXfv3ujdu7fp78jISJw8eRLLly83JTVr1qyBXq/Hhx9+CK1Wi5YtW+LQoUNYuHChokkNLz8RERHVYtnZ2WZLYaHcTSKVkZWVBR8fH9PfSUlJ6Nq1K7RarWldr169cPLkSdy8edPm5VcWkxoiIiIFiP/d/WTpIv7XpyY0NBSenp6mJTEx0ab1PHPmDN555x38/e9/N61LT09HYGCgWVzJ3+np6TYtXwYvPxERESnAIG4v1mwPAKmpqfDw+GvIEp1OV2b8jBkz8MYbb9xzn3/88QeaNfvr9v9Lly6hd+/eeOyxxzB27FjLK1tNmNQQERHVYh4eHmZJTXmee+45jBo16p4xkZGRpv9fvnwZ3bt3R+fOnfHuu++axQUFBSEjI8NsXcnfQUHyY3/ZCpMaIiIiBVT34Hv+/v7w9/evVOylS5fQvXt3REdHY9WqVVCrzXurxMTE4KWXXkJRUREcHR0BAJs3b0bTpk3h7V31g1WWh31qiIiIFFBy95M1S1W4dOkSunXrhrCwMLz11lu4evUq0tPTzfrKDBs2DFqtFmPGjMGxY8ewdu1avP3225g2bVqV1Kmy2FJDREREJps3b8aZM2dw5swZ1K9f3+yxks7Jnp6e2LRpE+Lj4xEdHQ0/Pz/MnDlT0du5gTqU1Czf/Se0LnLzFMmY0T2y4qA7uOucpcvILTRIxTs5yM2hY4n1v6dJxcc1D5AuQ3b+Gd/Cq1LxBW6BFQfdJeLkTqn4rL/L342Qv0FudE5HP7nnoenQRyoeAISzp1R8cqr8fGLfH8uoOOgOc+MaScXfeGK2VDwAuB/ZIBWvbvmAdBlGNz+peE223B0mklOiAQCKJFsCdp3NlC5D5yB3wSDM10UqPuVqnlR8VbV+lKWmzv00atSoCvveAEBUVBR++eWXKqmDpRS9/LRr1y488sgjCAkJgUqlwtdff232uBACM2fORHBwMJydnREbG4vTp08rU1kiIiIbKrn7yZqFzCma1OTl5aFNmzZYunRpmY8vWLAAS5YswYoVK7B37164urqiV69eKCgoqOaaEhERUU2n6OWnPn36oE+fspvAhRBYvHgxXn75ZfTv3x8A8PHHHyMwMBBff/01Hn/88eqsKhERkU3V1MtPtVmNvfspJSUF6enpiI2NNa3z9PREp06dkJSUVO52hYWFpYaMJiIiqmmMRmH1QuZqbFJTcutYWcMw32sI5sTERLPhokNDQ6u0nkRERJYwWtmfhjlNaTU2qbFUQkICsrKyTEtqaqrSVSIiIqJqUGNv6S4ZZjkjIwPBwcGm9RkZGWjbtm252+l0unLnvSAiIqop2KfG9mpsS01ERASCgoKwdetW07rs7Gzs3bsXMTExCtaMiIjIegYhrF7InKItNbm5uThz5ozp75SUFBw6dAg+Pj4ICwvDlClT8Nprr6Fx48aIiIjAK6+8gpCQEAwYMEC5ShMREVGNpGhSs3//fnTv3t30d8mcESNHjsTq1avxwgsvIC8vD+PGjUNmZibuv/9+/PTTT3ByclKqykRERDZh7R1MvPupNEWTmm7dupnmkSiLSqXC3LlzMXfu3GqslWXmbz8nFd+zaeVmSr2To0buamHqjXzpMqrapWz5gRNdHOWme6jvJ5f06vQ5UvEAYCyQG3rd+Ye3pctQy0570D5OKl7o5KcNUd3KkorvGWjJR4z8tBVSe3eVr5NDRCupeKOQm9oDkD+2Bo8gqfiZP52pOMhKozuGSW+TXyQ3/cuRi3LHSZZeL1cfaxhg3ajA1VfT2qPG9qkhIiIiklFj734iIiKyZ7z7yfaY1BARESnA2juYePdTabz8RERERHaBLTVEREQKMBoFDLz7yaaY1BARESnAYGVSY8229opJDRERkQKY1Nge+9QQERGRXWBLDRERkQIMRutaWwzyYzzaPSY1RERECuDlJ9vj5SciIiKyC2ypsZFOET5S8bLzOAHAH1fk5ijKzC+SLqOqtQv2kN6miY/cXE6qE1uk4osunJKKBwBtpNxcQKe+2CFdRvGtYqn48IfknofvI0Ol4gEgOyhKKt4tM0W6jJ4B3lLxFwrkZsBZvEu+TrIflf1ays+rpXOQ/EzIlJtHrX24l9z+AXg7OUrFN5ecdw2oeXM/VSe21NgekxoiIiIFcJwa2+PlJyIiIrILbKkhIiJSgEFYefmJcz+VwqSGiIhIAexTY3u8/ERERER2gS01RERECmBLje0xqSEiIlJAsVFAY0ViUsykphQmNURERApgS43tsU8NERER2QW21BARESmAg+/ZHpMaG7lwM18qXucgP4x6PQ+5Icj1xXJTuKZJDrtuiUgvnfQ2rtBLxWt8gqTi9WcOS8UDgMpbrozju1Kly7ggOc1FD8kPOPfmzaXiAcCpXhupeJVBbqoHAFDfkhsGP9BbblqF6tDMz0V6m1PXb0nFH07Plor/85rcZxQAtK7vKRW/7vhV6TLSsqv+c6emMghh1VgzHKemNF5+IiIiIrvAlhoiIiIFsKOw7TGpISIiUgCTGtvj5SciIiKyC2ypISIiUgBbamyPSQ0REZECDMIIg1HuLtW7tydzvPxEREREdoFJDRERkQJKBt+zdKmOwfcKCwvRtm1bqFQqHDp0yOyxw4cP44EHHoCTkxNCQ0OxYMGCKq9PRZjUEBERKcCahMba/jiV9cILLyAkJKTU+uzsbMTFxSE8PBzJycl48803MXv2bLz77rtVXqd7YZ8aIiIiBRQbAZVVs3TbsDJl+PHHH7Fp0yZ89dVX+PHHH80eW7NmDfR6PT788ENotVq0bNkShw4dwsKFCzFu3Liqrdg9sKWGiIioFsvOzjZbCgsLrd5nRkYGxo4di08++QQuLqWn/UhKSkLXrl2h1WpN63r16oWTJ0/i5s2bVpdvKbbU2IjsvEkx4fLz1Tg5yOWgbjq505slOdcQAOTrDVLxuUXyPy08JaeLMupcpeIdgiPkCgCgMsjNRxXR2l+6DMPhK1Lxzt5yc4MVX0+XigcASP6o1Ojk5zhTGeXmi8qTfE39vXO4VDwAOKpVUvFnb8rPZ/TdUQvOhwQXrUZ6G43c00aBBU0HsnPU2RODUUBtg1u6Q0NDzdbPmjULs2fPtni/QgiMGjUK48ePR4cOHXD+/PlSMenp6YiIMP/sDAwMND3mrdCcbExqiIiIFGCrpCY1NRUeHh6m9Tpd2b8EZ8yYgTfeeOOe+/zjjz+wadMm5OTkICEhweK6KYVJDRERUS3m4eFhltSU57nnnsOoUaPuGRMZGYlt27YhKSmpVHLUoUMHDB8+HB999BGCgoKQkZFh9njJ30FBQXJPwIaY1BARESnAVi01leXv7w9//4ovhS9ZsgSvvfaa6e/Lly+jV69eWLt2LTp16gQAiImJwUsvvYSioiI4OjoCADZv3oymTZsqdukJYFJDRESkCKOVt2VX1Tg1YWFhZn+7ud3uG9ewYUPUr18fADBs2DDMmTMHY8aMwfTp03H06FG8/fbbWLRoUZXUqbKY1BAREZEUT09PbNq0CfHx8YiOjoafnx9mzpyp6O3cAJMaIiIiRRiMwqpxaqprQssGDRpAiNJlRUVF4ZdffqmWOlQWkxoiIiIFCCEgrEhMyko06joOvkdERER2gS01RERECjBaOSlldUxoWdswqSEiIlKAEMKqS0i8/FQakxob0UgOo+6olr/y56at2tMV4S83vQAAHLuULRW/cGeKdBmy3u4g90ZXRXWXLyTzslT4fYteki4iOvuGVLwxR26+FX1GmlQ8ALhePSUVf0AVVnHQXZwc5d4brpLTJDTw1FYcdJcig9xrqqgG/oIe0Fp+QLS0HLk5hP7MvCVdRqcwuTFNtp66KhUvO5VLdRJGK/vU1MDXmdLYp4aIiIjsQo1OambPng2VSmW2NGvWTOlqERERWa2kT401C5mr8ZefWrZsiS1btpj+dnCo8VUmIiKqkDDeXqzZnszV+AzBwcFB0cmxiIiIqHao0ZefAOD06dMICQlBZGQkhg8fjgsXLtwzvrCwENnZ2WYLERFRTVNy95M1C5mr0UlNp06dsHr1avz0009Yvnw5UlJS8MADDyAnJ6fcbRITE+Hp6WlaQkNDq7HGRERElcM+NbZXo5OaPn364LHHHkNUVBR69eqFH374AZmZmfjiiy/K3SYhIQFZWVmmJTU1tRprTEREREqp8X1q7uTl5YUmTZrgzJkz5cbodDrodLpqrBUREZE8jlNjezW6peZuubm5OHv2LIKDg5WuChERkXX+l9RYuoBJTSk1Oqn55z//iZ07d+L8+fPYvXs3Bg4cCI1GgyeeeELpqhEREVENU6MvP128eBFPPPEErl+/Dn9/f9x///3Ys2cP/P39la4aERGRVYxCQGXFHUxG3v1USo1Oaj7//HOlq1BpBslmwPwi+flIPJ3kTpfs/FI6hxrdcFdpk/fLzcO18EH5EayKfeTmNCpO3iZdhuFWvlS8o1+gVLxr7GNS8QCQ5dVQKj6kWP5DN3HbWan4l2Pl6uQo9/IAABgl53bTqOQLGdRG7rL6+t/l5u7yd5Gf8yrYTW6bXg3l5nECAMdiufmi9l6Q+xys0XM/CSv71DCpKaVGJzVERET2ih2Fbc8+fpoTERFRnceWGiIiIgUYjYDKitYWI+d+KoVJDRERkQKsneqAfWpK4+UnIiIisgtsqSEiIlKAMN5erNmezDGpISIiUoDRKKzsU8PLT3fj5SciIiKyC2ypISIiUgDHqbE9JjVEREQKYFJje7z8RERERHaBLTUKySoslt7GXaeRii8olusan2NBnWqihT2DpOIdrpyWLkPo3KTiNS07SpcBo9ycNSoXT6l4YcHIXel5cq8RT8nXLAD0biE3h1V+kdzzkI0HAL1B7hdxem6hdBk1kYfk+dPqc6TL0GvdpeJzC+zjcwrghJZVgUkNERGRAnj5yfaY1BARESmAs3TbHvvUEBERkV1gSw0REZEChFFYNYAeLz+VxqSGiIhIAZzQ0vZ4+YmIiIjsAltqiIiIFMC7n2yPSQ0REZECjEYBcEJLm+LlJyIiIrILbKkhIiJSgDAaICRHDr97ezLHpEYhu05fk97m/ka+UvG5ktMe/HktXyreEi/1bCi9zetbz0rFa9JPyhWglm+wLNj9nVT8jWMp0mVonLRS8d5tW0nFZ94/SioeAJpmnZHb4Jb8h2638OZS8TN+lJ/mQlb7cC+p+NYBckP/A4CHk9yUBMFeTlLxf2YVSMUDgJeTq1T8LUf5552eWyQV3zLEQyr+t/M3peKrE5Ma2+PlJyIiIrILTGqIiIgUIIxGU2uNZYv85Kwyvv/+e3Tq1AnOzs7w9vbGgAEDzB6/cOEC+vXrBxcXFwQEBOD5559HcbGyE47y8hMREZEChMEAYbDi8pMV21bkq6++wtixYzFv3jz06NEDxcXFOHr0qOlxg8GAfv36ISgoCLt370ZaWhpGjBgBR0dHzJs3r8rqVREmNURERAoQwso+NaJqkpri4mJMnjwZb775JsaMGWNa36JFC9P/N23ahOPHj2PLli0IDAxE27Zt8eqrr2L69OmYPXs2tFq5PoG2wstPREREtVh2drbZUlhYaNX+Dhw4gEuXLkGtVqNdu3YIDg5Gnz59zFpqkpKS0Lp1awQGBprW9erVC9nZ2Th27JhV5VuDSQ0REZECrOtP81crT2hoKDw9PU1LYmKiVfU6d+4cAGD27Nl4+eWXsXHjRnh7e6Nbt264ceMGACA9Pd0soQFg+js9Pd2q8q3BpIaIiEgBtkpqUlNTkZWVZVoSEhLKLG/GjBlQqVT3XE6cOAHj/zogv/TSSxg8eDCio6OxatUqqFQqrFu3rtqOjyXYp4aIiKgW8/DwgIdHxeP3PPfccxg1atQ9YyIjI5GWlgbAvA+NTqdDZGQkLly4AAAICgrCvn37zLbNyMgwPaYUJjVEREQKqO7B9/z9/eHv719hXHR0NHQ6HU6ePIn7778fAFBUVITz588jPDwcABATE4PXX38dV65cQUBAAABg8+bN8PDwMEuGqhuTGiIiIgWUjFNjzfZVwcPDA+PHj8esWbMQGhqK8PBwvPnmmwCAxx57DAAQFxeHFi1a4KmnnsKCBQuQnp6Ol19+GfHx8dDpdFVSr8pgUkNERERm3nzzTTg4OOCpp57CrVu30KlTJ2zbtg3e3t4AAI1Gg40bN2LChAmIiYmBq6srRo4ciblz5ypabyY1tcg+yTlMHNQqqXh9cdWOTgkAfvqr0tvMiWskFW84+qNUvMjLlooHgMxTqVLxV36/IF2GzkNunAf3sMCKg+7ga8ySigcAcSNNKl7lLDd3EAA43TgnvY2M57tFSm8TmnVCKr44dY90Gca8HKn4GU07SMVP3i8/95Nbh/pS8UUGIV3G6Rt5UvEXb96SLqOmMhoNgBUtNcYqnPvJ0dERb731Ft56661yY8LDw/HDDz9UWR0swaSGiIhIAZzQ0vZ4SzcRERHZBbbUEBERKYAtNbbHpIaIiEgJBgOE2orEpAontKytmNQQEREpQAjrOgpX1YSWtRn71BAREZFdYEsNERGRAoTRaF1LTRUNvlebMakhIiJSgLBynBp2FC6Nl5+IiIjILrClhoiISAG3Lz9ZfgmJl59KY1Jjx7QOcg1x+Xr5pswXussNOV/kLP+S87p2Wq6MnEypeIfAUKl4AMhPvy4VbyiSP7ZO3nJTDDgGyT0PlaFYKh4AjLmZUvEarwDpMqCSe90ObRckFV8/+5RUPAAYXH2l4lXay9JlOLh4SMUXpxyVih8QFScVDwDX8ouk4osM8l+yLo4aqfgr2YXSZdRUvPxke7z8RERERHahViQ1S5cuRYMGDeDk5IROnTph3759SleJiIjIKiUjCluzkLkan9SsXbsW06ZNw6xZs3DgwAG0adMGvXr1wpUrV5SuGhERkcWMRoPVC5mr8UnNwoULMXbsWIwePRotWrTAihUr4OLigg8//FDpqhEREVENUqM7Cuv1eiQnJyMhIcG0Tq1WIzY2FklJSWVuU1hYiMLCvzqSZWdnV3k9iYiIZAmDEVBZ0VHYgo7Z9q5Gt9Rcu3YNBoMBgYGBZusDAwORnp5e5jaJiYnw9PQ0LaGh8ne2EBERVTUhrOxTw7mfSqnRSY0lEhISkJWVZVpSU1OVrhIREVEp7ChsezX68pOfnx80Gg0yMjLM1mdkZCAoqOyxKXQ6HXQ6XXVUj4iIiGqQGt1So9VqER0dja1bt5rWGY1GbN26FTExMQrWjIiIyDpsqbG9Gt1SAwDTpk3DyJEj0aFDB3Ts2BGLFy9GXl4eRo8erXTViIiILCaMBus6CjOpKaXGJzVDhw7F1atXMXPmTKSnp6Nt27b46aefSnUeLo8QAgCgv5VXldWsFkJy2gNNsdzw43rJIdEBIEfy7jJtkfxLTpeTKxVflHdLKt4hV/61kauXO1a5xfJTEjhKlpEt+bw1OTlS8QAg8vIly5A7dwAgtHIf1HmFzlLx2Sr5OhkNkq/bXLnjBAAqB7nzbZA8F3m58ue7sFju7hpLpkkolNxGny9//qT2/7/vipLvjiplKIJVpRjkP7PtnUpUy5lTzsWLF3kHFBERSUlNTUX9+vWrZN8FBQWIiIgo9y5eGUFBQUhJSYGTk5MNalb72X1SYzQacfnyZbi7u0OlUpnWZ2dnIzQ0FKmpqfDwkJtIrjbj8+bzrgv4vPm8LSWEQE5ODkJCQqBWV12304KCAuj1eqv3o9VqmdDcocZffrKWWq2+Z7bt4eFRp978Jfi86xY+77qFz9s6np6eNqjNvTk5OTEZqQI1+u4nIiIiospiUkNERER2oc4mNTqdDrNmzapzA/XxefN51wV83nzeVDfZfUdhIiIiqhvqbEsNERER2RcmNURERGQXmNQQERGRXWBSQ0RERHahziY1S5cuRYMGDeDk5IROnTph3759SlepSs2ePRsqlcpsadasmdLVsrldu3bhkUceQUhICFQqFb7++muzx4UQmDlzJoKDg+Hs7IzY2FicPn1amcraUEXPe9SoUaXOf+/evZWprI0kJibivvvug7u7OwICAjBgwACcPHnSLKagoADx8fHw9fWFm5sbBg8ejIyMDIVqbBuVed7dunUrdb7Hjx+vUI1tY/ny5YiKijINsBcTE4Mff/zR9Lg9nmuSVyeTmrVr12LatGmYNWsWDhw4gDZt2qBXr164cuWK0lWrUi1btkRaWppp+fXXX5Wuks3l5eWhTZs2WLp0aZmPL1iwAEuWLMGKFSuwd+9euLq6olevXigoKKjmmtpWRc8bAHr37m12/j/77LNqrKHt7dy5E/Hx8dizZw82b96MoqIixMXFIS/vrwlKp06diu+++w7r1q3Dzp07cfnyZQwaNEjBWluvMs8bAMaOHWt2vhcsWKBQjW2jfv36mD9/PpKTk7F//3706NED/fv3x7FjxwDY57kmC4g6qGPHjiI+Pt70t8FgECEhISIxMVHBWlWtWbNmiTZt2ihdjWoFQGzYsMH0t9FoFEFBQeLNN980rcvMzBQ6nU589tlnCtSwatz9vIUQYuTIkaJ///6K1Ke6XLlyRQAQO3fuFELcPreOjo5i3bp1ppg//vhDABBJSUlKVdPm7n7eQgjx4IMPismTJytXqWri7e0t3n///Tpzrqlida6lRq/XIzk5GbGxsaZ1arUasbGxSEpKUrBmVe/06dMICQlBZGQkhg8fjgsXLihdpWqVkpKC9PR0s3Pv6emJTp062f25B4AdO3YgICAATZs2xYQJE3D9+nWlq2RTWVlZAAAfHx8AQHJyMoqKiszOd7NmzRAWFmZX5/vu511izZo18PPzQ6tWrZCQkID8/HwlqlclDAYDPv/8c+Tl5SEmJqbOnGuqmN1PaHm3a9euwWAwIDAw0Gx9YGAgTpw4oVCtql6nTp2wevVqNG3aFGlpaZgzZw4eeOABHD16FO7u7kpXr1qkp6cDQJnnvuQxe9W7d28MGjQIEREROHv2LF588UX06dMHSUlJ0Gg0SlfPakajEVOmTEGXLl3QqlUrALfPt1arhZeXl1msPZ3vsp43AAwbNgzh4eEICQnB4cOHMX36dJw8eRLr169XsLbWO3LkCGJiYlBQUAA3Nzds2LABLVq0wKFDh+z+XFPl1Lmkpq7q06eP6f9RUVHo1KkTwsPD8cUXX2DMmDEK1oyqw+OPP276f+vWrREVFYWGDRtix44d6Nmzp4I1s434+HgcPXrULvuJ3Ut5z3vcuHGm/7du3RrBwcHo2bMnzp49i4YNG1Z3NW2madOmOHToELKysvDll19i5MiR2Llzp9LVohqkzl1+8vPzg0ajKdUrPiMjA0FBQQrVqvp5eXmhSZMmOHPmjNJVqTYl57eun3sAiIyMhJ+fn12c/4kTJ2Ljxo3Yvn076tevb1ofFBQEvV6PzMxMs3h7Od/lPe+ydOrUCQBq/fnWarVo1KgRoqOjkZiYiDZt2uDtt9+2+3NNlVfnkhqtVovo6Ghs3brVtM5oNGLr1q2IiYlRsGbVKzc3F2fPnkVwcLDSVak2ERERCAoKMjv32dnZ2Lt3b5069wBw8eJFXL9+vVaffyEEJk6ciA0bNmDbtm2IiIgwezw6OhqOjo5m5/vkyZO4cOFCrT7fFT3vshw6dAgAavX5LovRaERhYaHdnmuygNI9lZXw+eefC51OJ1avXi2OHz8uxo0bJ7y8vER6errSVasyzz33nNixY4dISUkR//3vf0VsbKzw8/MTV65cUbpqNpWTkyMOHjwoDh48KACIhQsXioMHD4o///xTCCHE/PnzhZeXl/jmm2/E4cOHRf/+/UVERIS4deuWwjW3zr2ed05OjvjnP/8pkpKSREpKitiyZYto3769aNy4sSgoKFC66habMGGC8PT0FDt27BBpaWmmJT8/3xQzfvx4ERYWJrZt2yb2798vYmJiRExMjIK1tl5Fz/vMmTNi7ty5Yv/+/SIlJUV88803IjIyUnTt2lXhmltnxowZYufOnSIlJUUcPnxYzJgxQ6hUKrFp0yYhhH2ea5JXJ5MaIYR45513RFhYmNBqtaJjx45iz549SlepSg0dOlQEBwcLrVYr6tWrJ4YOHSrOnDmjdLVsbvv27QJAqWXkyJFCiNu3db/yyisiMDBQ6HQ60bNnT3Hy5EllK20D93re+fn5Ii4uTvj7+wtHR0cRHh4uxo4dW+uT+LKeLwCxatUqU8ytW7fEs88+K7y9vYWLi4sYOHCgSEtLU67SNlDR875w4YLo2rWr8PHxETqdTjRq1Eg8//zzIisrS9mKW+npp58W4eHhQqvVCn9/f9GzZ09TQiOEfZ5rkqcSQojqaxciIiIiqhp1rk8NERER2ScmNURERGQXmNQQERGRXWBSQ0RERHaBSQ0RERHZBSY1REREZBeY1BAREZFdYFJDREREdoFJDREREdkFJjVERERkF5jUENUhV69eRVBQEObNm2dat3v3bmi1WrMZjomIaiPO/URUx/zwww8YMGAAdu/ejaZNm6Jt27bo378/Fi5cqHTViIiswqSGqA6Kj4/Hli1b0KFDBxw5cgS//fYbdDqd0tUiIrIKkxqiOujWrVto1aoVUlNTkZycjNatWytdJSIiq7FPDVEddPbsWVy+fBlGoxHnz59XujpERDbBlhqiOkav16Njx45o27YtmjZtisWLF+PIkSMICAhQumpERFZhUkNUxzz//PP48ssv8fvvv8PNzQ0PPvggPD09sXHjRqWrRkRkFV5+IqpDduzYgcWLF+OTTz6Bh4cH1Go1PvnkE/zyyy9Yvny50tUjIrIKW2qIiIjILrClhoiIiOwCkxoiIiKyC0xqiIiIyC4wqSEiIiK7wKSGiIiI7AKTGiIiIrILTGqIiIjILjCpISIiIrvApIaIiIjsApMaIiIisgtMaoiIiMguMKkhIiIiu/D/lif+JS9l+yIAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_feats_standardized.isel(feature=0,starttime=15).plot()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "w2T48h8ykonf",
        "outputId": "a751198e-f5b8-4083-c511-a6d5e6d4741d"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.QuadMesh at 0x7f0a94af4ca0>"
            ]
          },
          "metadata": {},
          "execution_count": 21
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_feats.isel(feature=1,starttime=100).plot()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "PJh6t50olRlG",
        "outputId": "2c6ef7a8-8dc5-4e70-9659-932b131201d6"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.QuadMesh at 0x7f0a949c3430>"
            ]
          },
          "metadata": {},
          "execution_count": 22
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_feats_standardized.isel(feature=1,starttime=100).plot()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "EO_ugNfykwE2",
        "outputId": "9789ba22-c724-404b-e291-cdd11435bc43"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.QuadMesh at 0x7f0a94893190>"
            ]
          },
          "metadata": {},
          "execution_count": 23
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's proceed to build a simple convolutional neural network that will take images as the input and predict the respective storm's maximum updraft helicity."
      ],
      "metadata": {
        "id": "yS-guBx7gOkQ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import keras\n",
        "from keras import layers"
      ],
      "metadata": {
        "id": "KDM-f2a1fnlB"
      },
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# lets use two input variables (dbz_grid and ctt_grid)\n",
        "number_of_features = 2"
      ],
      "metadata": {
        "id": "94Cj7eVlhByj"
      },
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = keras.Sequential()\n",
        "\n",
        "# input\n",
        "model.add(keras.Input(shape=(32, 32, number_of_features)))  # 32x32 images\n",
        "\n",
        "# data augmentation layers\n",
        "#model.add(layers.RandomFlip(\"horizontal_and_vertical\"))\n",
        "#model.add(layers.RandomRotation(factor=(-0.5, 0.5)))\n",
        "\n",
        "# hidden layers\n",
        "model.add(layers.Conv2D(32, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.MaxPooling2D(2))\n",
        "model.add(layers.Dropout(0.2))\n",
        "\n",
        "model.add(layers.Conv2D(64, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.MaxPooling2D(2))\n",
        "model.add(layers.Dropout(0.2))\n",
        "\n",
        "model.add(layers.Conv2D(128, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.MaxPooling2D(2))\n",
        "model.add(layers.Dropout(0.2))\n",
        "\n",
        "# Now that we have 4x4 feature maps, now apply global max pooling.\n",
        "model.add(layers.GlobalMaxPooling2D())\n",
        "\n",
        "# Finally, we add last fully connected (dense) layers\n",
        "model.add(layers.Dense(64))\n",
        "model.add(layers.Dense(1, activation=\"linear\"))\n",
        "\n",
        "# print model summary\n",
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aGZqhCVfII4J",
        "outputId": "37b7ad25-596b-4f1c-d652-4037031d8431"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d (Conv2D)             (None, 32, 32, 32)        608       \n",
            "                                                                 \n",
            " max_pooling2d (MaxPooling2  (None, 16, 16, 32)        0         \n",
            " D)                                                              \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 16, 16, 32)        0         \n",
            "                                                                 \n",
            " conv2d_1 (Conv2D)           (None, 16, 16, 64)        18496     \n",
            "                                                                 \n",
            " max_pooling2d_1 (MaxPoolin  (None, 8, 8, 64)          0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 8, 8, 64)          0         \n",
            "                                                                 \n",
            " conv2d_2 (Conv2D)           (None, 8, 8, 128)         73856     \n",
            "                                                                 \n",
            " max_pooling2d_2 (MaxPoolin  (None, 4, 4, 128)         0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_2 (Dropout)         (None, 4, 4, 128)         0         \n",
            "                                                                 \n",
            " global_max_pooling2d (Glob  (None, 128)               0         \n",
            " alMaxPooling2D)                                                 \n",
            "                                                                 \n",
            " dense (Dense)               (None, 64)                8256      \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 1)                 65        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 101281 (395.63 KB)\n",
            "Trainable params: 101281 (395.63 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Proceeding with compiling our model and selecting some metrics to use."
      ],
      "metadata": {
        "id": "AXkjSNM3mD2u"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Some common evaluation metrics used with regression problems are:\n",
        "\n",
        "Mean absolute error (MAE): $\\frac{1}{N} \\sum\\limits_{i = 1}^{N} \\lvert \\hat{y}_i - y_i \\rvert$\n",
        "\n",
        "Mean squared error (MSE): $\\frac{1}{N} \\sum\\limits_{i = 1}^{N} (\\hat{y}_i - y_i)^2$\n",
        "\n",
        "\n",
        "More evalution metrics can be found [here](https://scikit-learn.org/stable/modules/model_evaluation.html)."
      ],
      "metadata": {
        "id": "knrWFwJ5LU2A"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "METRICS = [\n",
        "    keras.metrics.MeanSquaredError(name='mean_squared_error'),\n",
        "    keras.metrics.RootMeanSquaredError(name='root_mean_squared_error'),\n",
        "]"
      ],
      "metadata": {
        "id": "lPb4PhGRlZgN"
      },
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(\n",
        "    optimizer=keras.optimizers.Adam(learning_rate=0.001),\n",
        "    loss=\"mean_squared_error\",\n",
        "    metrics=METRICS\n",
        ")"
      ],
      "metadata": {
        "id": "o82dwXizjNTa"
      },
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "BATCH_SIZE = 32\n",
        "EPOCHS = 10"
      ],
      "metadata": {
        "id": "QEdyoR97MTrv"
      },
      "execution_count": 29,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "history = model.fit(\n",
        "    x=train_feats_standardized.values,\n",
        "    y=train_label_standardized.values,\n",
        "    batch_size=BATCH_SIZE,\n",
        "    epochs=EPOCHS,\n",
        "    validation_split=0.1,\n",
        "    shuffle=True,\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QfhYCYuVmXu7",
        "outputId": "64b9ec85-c52e-4af2-b77f-bd7d300df2b3"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "44/44 [==============================] - 3s 50ms/step - loss: 0.5513 - mean_squared_error: 0.5513 - root_mean_squared_error: 0.7425 - val_loss: 0.5110 - val_mean_squared_error: 0.5110 - val_root_mean_squared_error: 0.7149\n",
            "Epoch 2/10\n",
            "44/44 [==============================] - 2s 55ms/step - loss: 0.3912 - mean_squared_error: 0.3912 - root_mean_squared_error: 0.6254 - val_loss: 0.3774 - val_mean_squared_error: 0.3774 - val_root_mean_squared_error: 0.6144\n",
            "Epoch 3/10\n",
            "44/44 [==============================] - 2s 45ms/step - loss: 0.2830 - mean_squared_error: 0.2830 - root_mean_squared_error: 0.5319 - val_loss: 0.3459 - val_mean_squared_error: 0.3459 - val_root_mean_squared_error: 0.5881\n",
            "Epoch 4/10\n",
            "44/44 [==============================] - 2s 45ms/step - loss: 0.2016 - mean_squared_error: 0.2016 - root_mean_squared_error: 0.4490 - val_loss: 0.2362 - val_mean_squared_error: 0.2362 - val_root_mean_squared_error: 0.4860\n",
            "Epoch 5/10\n",
            "44/44 [==============================] - 2s 45ms/step - loss: 0.1523 - mean_squared_error: 0.1523 - root_mean_squared_error: 0.3902 - val_loss: 0.2305 - val_mean_squared_error: 0.2305 - val_root_mean_squared_error: 0.4801\n",
            "Epoch 6/10\n",
            "44/44 [==============================] - 2s 46ms/step - loss: 0.1418 - mean_squared_error: 0.1418 - root_mean_squared_error: 0.3766 - val_loss: 0.2874 - val_mean_squared_error: 0.2874 - val_root_mean_squared_error: 0.5361\n",
            "Epoch 7/10\n",
            "44/44 [==============================] - 2s 47ms/step - loss: 0.1236 - mean_squared_error: 0.1236 - root_mean_squared_error: 0.3516 - val_loss: 0.2919 - val_mean_squared_error: 0.2919 - val_root_mean_squared_error: 0.5403\n",
            "Epoch 8/10\n",
            "44/44 [==============================] - 2s 51ms/step - loss: 0.1109 - mean_squared_error: 0.1109 - root_mean_squared_error: 0.3331 - val_loss: 0.3426 - val_mean_squared_error: 0.3426 - val_root_mean_squared_error: 0.5853\n",
            "Epoch 9/10\n",
            "44/44 [==============================] - 2s 44ms/step - loss: 0.1140 - mean_squared_error: 0.1140 - root_mean_squared_error: 0.3377 - val_loss: 0.2217 - val_mean_squared_error: 0.2217 - val_root_mean_squared_error: 0.4708\n",
            "Epoch 10/10\n",
            "44/44 [==============================] - 2s 44ms/step - loss: 0.1058 - mean_squared_error: 0.1058 - root_mean_squared_error: 0.3253 - val_loss: 0.3122 - val_mean_squared_error: 0.3122 - val_root_mean_squared_error: 0.5587\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history[\"mean_squared_error\"], c='k')\n",
        "plt.plot(history.history[\"val_mean_squared_error\"], c='r')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "KD2-89DcoW44",
        "outputId": "e20c4f76-8b3a-4c41-bfa8-d0e7b8584ebb"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Lets now evaluate the model using the test dataset. We need to standardize the data, but using the mean and standard deviation of the training set."
      ],
      "metadata": {
        "id": "rwxlG2G0xcQ6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def z_score_testing(da, testdata):\n",
        "    \"\"\"\n",
        "    Compute z-score for testing data.\n",
        "    Args:\n",
        "        da (array): Array of training variable.\n",
        "    \"\"\"\n",
        "    return (testdata - np.mean(da)) / np.std(da)"
      ],
      "metadata": {
        "id": "7zaFmAIUorvC"
      },
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# normalization of our training data (note each feature individually processed!)\n",
        "test_feats_standardized = xr.concat([\n",
        "    z_score_testing(\n",
        "        train_feats.isel(feature=0).expand_dims(dim='feature', axis=3),\n",
        "        test_feats.isel(feature=0).expand_dims(dim='feature', axis=3)\n",
        "        ),\n",
        "    z_score_testing(\n",
        "        train_feats.isel(feature=1).expand_dims(dim='feature', axis=3),\n",
        "        test_feats.isel(feature=1).expand_dims(dim='feature', axis=3)\n",
        "        )],\n",
        "                                     dim='feature')\n",
        "\n",
        "test_label_standardized = z_score_testing(train_label, test_label)"
      ],
      "metadata": {
        "id": "3jR6rF-lx55_"
      },
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate the model on the test data using `evaluate`\n",
        "print(\"Evaluate on test data\")\n",
        "results = model.evaluate(test_feats_standardized.values, test_label_standardized.values, batch_size=128)\n",
        "print(results)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZXfumUFnor4-",
        "outputId": "68d07abb-0a03-41ee-d736-cba8d489fa52"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluate on test data\n",
            "6/6 [==============================] - 0s 41ms/step - loss: 0.3359 - mean_squared_error: 0.3359 - root_mean_squared_error: 0.5796\n",
            "[0.3358811140060425, 0.3358811140060425, 0.5795525312423706]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# generate predictions using the trained model\n",
        "predictions = model.predict(test_feats_standardized.values)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tSHfeMExzVqb",
        "outputId": "177331f3-7a57-4f3e-977c-5c6d22a5cb75"
      },
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "21/21 [==============================] - 0s 13ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# visualize predictions vs ground truth\n",
        "plt.scatter(predictions, test_label_standardized.values)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "xzOMgkXozb6V",
        "outputId": "252df329-f8d0-415a-c27a-8132cc0ebda1"
      },
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Tutorial Practice: How does batch size impact model training?\n",
        "\n",
        "---"
      ],
      "metadata": {
        "id": "6soWOXwjyo8t"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# recreate the model object"
      ],
      "metadata": {
        "id": "PZlR8mqqy3py"
      },
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# recompile the model"
      ],
      "metadata": {
        "id": "HOB_0mQEzJwz"
      },
      "execution_count": 38,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# retrain the model"
      ],
      "metadata": {
        "id": "5oNHXJ4_zJ3I"
      },
      "execution_count": 39,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# evaluate the new model's skill"
      ],
      "metadata": {
        "id": "rMHPapkbzJ9m"
      },
      "execution_count": 40,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Other sample questions for practice:\n",
        "\n",
        "*   How does the learning rate impact model training?\n",
        "*   Does data augmentation help with model skill?\n",
        "*   Does reducing regularization impact overfitting (hint: reduce dropout percentage)?"
      ],
      "metadata": {
        "id": "hsqlwFRY1aXd"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# more cells for practicing"
      ],
      "metadata": {
        "id": "8rhORSiS4Dz3"
      },
      "execution_count": 41,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# What happens if your model is not performing well? This may be related to issues in bias and/or variance...\n",
        "\n",
        "---\n",
        "\n",
        "### Bias / Variance Trade-off\n",
        "\n",
        "*Subsection on bias/variance trade-off sourced from the AMS 2022 AI Short Course and the following [paper](http://scott.fortmann-roe.com/docs/BiasVariance.html)*\n",
        "\n",
        "An important concept in machine learning is that of the \"bias/variance trade-off.\" Prediction errors from any model can be decomposed into several terms: 1) error due to bias, 2) error due to variance, and 3) irreducible errors or noise. This concept refers to the trade-off that occurs in a models ability to reduce error in the bias and variance terms.\n",
        "\n",
        "<img src=\"https://www.cs.cornell.edu/courses/cs4780/2018fa/lectures/images/bias_variance/bullseye.png\" height= \"600\">\n",
        "\n",
        "(image credit: [paper](http://scott.fortmann-roe.com/docs/BiasVariance.html))\n",
        "\n",
        "\n",
        "### Bias\n",
        "\n",
        "Bias is due to your model being consistently getting a certain type of error in relation to a particular kind of solution. For example, maximum temperatures in relation to the diurnal cycle might be predicted by your model consistently higher than what was observed. This error may be irreducible with more training data and be related to your model (i.e., inherent to your model). The error due to bias is taken as the difference between the expected (or average) prediction of our model and the correct value which we are trying to predict. Bias measures how far off in general these models' predictions are from the correct value.\n",
        "\n",
        "### Variance\n",
        "\n",
        "Variance relates to changes in your model's predictions based on new iterations of your model and thus captures how much your predictions change if your model is trained on a different training set (e.g., different data ordering). We can also think of this as how \"over-specialized\" your machine learning model may be to a particular training set (overfitting). The error due to variance is taken as the variability of a model's prediction for a given data point. Again, variance is how much the predictions for a given point vary between different realizations of the model.\n",
        "\n",
        "### Irreducible Error/Noise\n",
        "\n",
        "How big is the data-intrinsic noise? This error relates to ambiguity due to your data distribution and feature representation. You can never beat this, it is an aspect of the data.\n",
        "\n",
        "# So, what do we do with this information?\n",
        "\n",
        "If the training error is below your desired error threshold, but the test error is significantly higher, then the cause of the model's poor performance is high variance. In other words, we have overfit to the training set. We can remedy this issue by adding more training data and/or reducing the model's complexity. For decision trees, we can also perform \"bagging.\"\n",
        "\n",
        "If the training error and the test error are both above your desired error threshold, then the model being used is not robust or complex enough to produce accurate predictions. In other words, we have underfit to the training set. We can remedy this issue by using more complex model and/or adding more variables (features). For decision trees, we can also perform \"boosting.\"\n",
        "\n",
        "For more information, see [Cornell Lecture](https://www.cs.cornell.edu/courses/cs4780/2018fa/lectures/lecturenote12.html)."
      ],
      "metadata": {
        "id": "71KIzoONMT3P"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Classification\n",
        "\n",
        "---\n",
        "\n",
        "**In classification problems**, we aim to predict the class of *something*. Classes are discrete in nature. We can frame our problem to be binary classification (two classes; e.g., True or False, or Yes or No problems). Classification problems can also be multiclass in nature.\n",
        "\n",
        "Continuous data could be grouped or binned into certain classes, and thus, continuous data can be converted to classes (e.g., binning of continuous maximum tropical cyclone wind speeds into hurricane intensities based on the Saffir-Simpson Hurricane Wind Scale, e.g., category 1, category 2, etc.).\n",
        "\n",
        "Ways to convert continuous data into classes include **integer encoding or one-hot encoding**. Integer encoding involves assigning an integer value to each unique category, and this can work well for data that have an underlying natural ordered relationship between each other, since machine learning models may be able to understand and harness this relationship. For one-hot encoding, a binary variable (i.e., zero or one) is added at a certain position in each sample's vector, which is meant to represent the respective sample's class, e.g., `[0, 0, 1]` vs `[1, 0, 0]`."
      ],
      "metadata": {
        "id": "e3fmySjlPp8p"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![supervised](https://drive.google.com/uc?export=view&id=1jDl42c0CSN9dvp37TMWJ1qI6fOQMGp4e)\n"
      ],
      "metadata": {
        "id": "aBFaBbXNARjL"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Some examples of classification problems in the environmental sciences could include:\n",
        "\n",
        "*   whether a thunderstorm will produce severe hazards or be non-severe (i.e., binary) (e.g., Molina et al. 2021)\n",
        "*   whether a polynya event is occurring in the Ross Sea, Antarctica (i.e., binary) (e.g., DuVivier et al. 2023)\n",
        "*   type of cloud(s) that is(are) in a citizen science photograph (i.e., multiclass)\n",
        "*   subseasonal prediction of a North American weather regime class (i.e., multiclass) (e.g., Molina et al. 2023)\n",
        "\n",
        "Similar to how explained in the regression section, when using various input variables (i.e., features or channels), normalizing or standardizing the variables is needed to make sure they have a similar dynamic range."
      ],
      "metadata": {
        "id": "BrOXsxoUAjm7"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Supervised Learning Classification Example\n",
        "\n",
        "---"
      ],
      "metadata": {
        "id": "Hd9l4RhAC6gu"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Using our work from the previous section, we will now use the data and reframe our problem to be a classification task. As before, we will use two input features."
      ],
      "metadata": {
        "id": "CFMnVVa5zti6"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "However, we will use updraft helicity values (which represent storm updraft rotation) and convert this continuous value into two classes, \"not potentially severe\" and \"potentially severe\". More strongly rotating storms have a greater likelihood of producing severe hazards like tornadoes and large hail. We will use 75 m2/s2 as the delineation mark between the two classes. The two classes will be creating using integer encoding first, with 0 representing non-severe and 1 representing potentially severe, and then converting using one-hot encoding for subsequent training."
      ],
      "metadata": {
        "id": "y5R40_674lUy"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "features = xr.concat([\n",
        "    ds['dbz_grid'].expand_dims(dim='feature', axis=3),\n",
        "    ds['ctt_grid'].expand_dims(dim='feature', axis=3)],\n",
        "          dim='feature')\n",
        "\n",
        "labels = ds['uh25_grid'].expand_dims(dim='feature', axis=3).max('x').max('y')\n",
        "\n",
        "train_feats, train_label, test_feats, test_label = random_split(features, labels)\n",
        "\n",
        "# normalization of our training data (note each feature individually processed!)\n",
        "train_feats_standardized = xr.concat([\n",
        "    z_score(train_feats.isel(feature=0).expand_dims(dim='feature', axis=3)),\n",
        "    z_score(train_feats.isel(feature=1).expand_dims(dim='feature', axis=3))],\n",
        "                                     dim='feature')\n",
        "\n",
        "train_label_ints = xr.where(train_label >= 75, 1, 0)"
      ],
      "metadata": {
        "id": "hj6bAjZ7zt5v"
      },
      "execution_count": 42,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(train_label.values[-2:])\n",
        "print(train_label_ints.values[-2:])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dnlyxpuOzuCl",
        "outputId": "6c6019a7-0feb-4384-f81a-ed1228ec4146"
      },
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[162.55905  ]\n",
            " [  1.8748366]]\n",
            "[[1]\n",
            " [0]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# convert labels into classes using one-hot encoding\n",
        "onehot_layer = keras.layers.CategoryEncoding(num_tokens=2, output_mode=\"one_hot\")\n",
        "train_y = onehot_layer(train_label_ints.astype(int))"
      ],
      "metadata": {
        "id": "LzGNl7wUzuKz"
      },
      "execution_count": 44,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_y[-2:]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K7axDV9KzuY3",
        "outputId": "7f93d34f-81d1-42ba-a1c6-2f77b93e015d"
      },
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(2, 2), dtype=float32, numpy=\n",
              "array([[0., 1.],\n",
              "       [1., 0.]], dtype=float32)>"
            ]
          },
          "metadata": {},
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# out of more than 1000 samples, only 189 are the \"severe\" class\n",
        "print(train_y[:,1].numpy().sum())\n",
        "print(train_y[:,1].numpy().shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lsRwqJjG7_Ec",
        "outputId": "3fa03eee-cd3d-4ea4-8155-6cc421233736"
      },
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "189.0\n",
            "(1556,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Class imbalance is a major issue with a minority class that is a rare event, like severe hazards. To address this limitation, we create class weights that help the machine learning model learn that the minority class is important to get correct."
      ],
      "metadata": {
        "id": "12FrqE9C7_mS"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# generate class weights due to class imbalance issues\n",
        "# use the second class as the positive one (using 1 index)\n",
        "counts = np.bincount(train_y[:,1].numpy().astype(int))\n",
        "\n",
        "print(\"Number of positive samples in training data: {} ({:.2f}% of total)\".format(\n",
        "    counts[1], 100 * float(counts[1]) / len(train_y)))\n",
        "\n",
        "weight_for_0 = 1.0 / counts[0]\n",
        "weight_for_1 = 1.0 / counts[1]\n",
        "\n",
        "class_weight = {0: weight_for_0, 1: weight_for_1}\n",
        "print(class_weight)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "asYu-Bio70M6",
        "outputId": "b51e7a8d-b328-428d-bb5a-71eeb71dbb51"
      },
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of positive samples in training data: 189 (12.15% of total)\n",
            "{0: 0.000731528895391368, 1: 0.005291005291005291}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# metrics that are useful for classification, particularly for two classes\n",
        "METRICS = [\n",
        "    keras.metrics.BinaryCrossentropy(name='cross entropy'),  # same as model's loss\n",
        "    keras.metrics.TruePositives(name='tp'),\n",
        "    keras.metrics.FalsePositives(name='fp'),\n",
        "    keras.metrics.TrueNegatives(name='tn'),\n",
        "    keras.metrics.FalseNegatives(name='fn'),\n",
        "    keras.metrics.BinaryAccuracy(name='accuracy'),\n",
        "    keras.metrics.Precision(name='precision'),\n",
        "    keras.metrics.Recall(name='recall'),\n",
        "    keras.metrics.AUC(name='auc'),\n",
        "    keras.metrics.AUC(name='prc', curve='PR'), # precision-recall curve\n",
        "]"
      ],
      "metadata": {
        "id": "xHmN3l2z70QS"
      },
      "execution_count": 48,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = keras.Sequential()\n",
        "\n",
        "# input\n",
        "model.add(keras.Input(shape=(32, 32, number_of_features)))  # 32x32 images\n",
        "\n",
        "# data augmentation layers\n",
        "#model.add(layers.RandomFlip(\"horizontal_and_vertical\"))\n",
        "#model.add(layers.RandomRotation(factor=(-0.5,0.5)))\n",
        "\n",
        "# hidden layers\n",
        "model.add(layers.Conv2D(32, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.MaxPooling2D(2))\n",
        "model.add(layers.Dropout(0.2))\n",
        "\n",
        "model.add(layers.Conv2D(64, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.MaxPooling2D(2))\n",
        "model.add(layers.Dropout(0.2))\n",
        "\n",
        "model.add(layers.Conv2D(128, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.MaxPooling2D(2))\n",
        "model.add(layers.Dropout(0.2))\n",
        "\n",
        "# Now that we have 4x4 feature maps, now apply global max pooling.\n",
        "model.add(layers.GlobalMaxPooling2D())\n",
        "\n",
        "# Finally, we add last fully connected (dense) layers\n",
        "model.add(layers.Dense(64))\n",
        "model.add(layers.Dense(2, activation=\"softmax\"))\n",
        "\n",
        "# print model summary\n",
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BHBpYKnh70WR",
        "outputId": "96ef8716-269e-49ed-8ecb-f8b76c582b8b"
      },
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d_3 (Conv2D)           (None, 32, 32, 32)        608       \n",
            "                                                                 \n",
            " max_pooling2d_3 (MaxPoolin  (None, 16, 16, 32)        0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_3 (Dropout)         (None, 16, 16, 32)        0         \n",
            "                                                                 \n",
            " conv2d_4 (Conv2D)           (None, 16, 16, 64)        18496     \n",
            "                                                                 \n",
            " max_pooling2d_4 (MaxPoolin  (None, 8, 8, 64)          0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_4 (Dropout)         (None, 8, 8, 64)          0         \n",
            "                                                                 \n",
            " conv2d_5 (Conv2D)           (None, 8, 8, 128)         73856     \n",
            "                                                                 \n",
            " max_pooling2d_5 (MaxPoolin  (None, 4, 4, 128)         0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_5 (Dropout)         (None, 4, 4, 128)         0         \n",
            "                                                                 \n",
            " global_max_pooling2d_1 (Gl  (None, 128)               0         \n",
            " obalMaxPooling2D)                                               \n",
            "                                                                 \n",
            " dense_2 (Dense)             (None, 64)                8256      \n",
            "                                                                 \n",
            " dense_3 (Dense)             (None, 2)                 130       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 101346 (395.88 KB)\n",
            "Trainable params: 101346 (395.88 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(\n",
        "    optimizer=keras.optimizers.Adam(learning_rate=0.001),\n",
        "    loss=keras.losses.BinaryCrossentropy(), # categorical loss function\n",
        "    metrics=METRICS\n",
        ")"
      ],
      "metadata": {
        "id": "xS8iZ6Ro70bD"
      },
      "execution_count": 50,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "BATCH_SIZE = 32\n",
        "EPOCHS = 10"
      ],
      "metadata": {
        "id": "m0jmE_7k823F"
      },
      "execution_count": 51,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"Fit model on training data\")\n",
        "history = model.fit(\n",
        "    x=train_feats_standardized.values,\n",
        "    y=train_y,\n",
        "    batch_size=BATCH_SIZE,\n",
        "    epochs=EPOCHS,\n",
        "    validation_split=0.1,\n",
        "    shuffle=True,\n",
        "    class_weight=class_weight,\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "o3wCne8Q826e",
        "outputId": "46f513ad-af20-4a71-a49c-4d395053948f"
      },
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fit model on training data\n",
            "Epoch 1/10\n",
            "44/44 [==============================] - 5s 68ms/step - loss: 7.5593e-04 - cross entropy: 0.6303 - tp: 803.0000 - fp: 597.0000 - tn: 803.0000 - fn: 597.0000 - accuracy: 0.5736 - precision: 0.5736 - recall: 0.5736 - auc: 0.6796 - prc: 0.7109 - val_loss: 0.7091 - val_cross entropy: 0.7091 - val_tp: 69.0000 - val_fp: 87.0000 - val_tn: 69.0000 - val_fn: 87.0000 - val_accuracy: 0.4423 - val_precision: 0.4423 - val_recall: 0.4423 - val_auc: 0.6096 - val_prc: 0.6809\n",
            "Epoch 2/10\n",
            "44/44 [==============================] - 2s 48ms/step - loss: 6.0370e-04 - cross entropy: 0.5456 - tp: 876.0000 - fp: 524.0000 - tn: 876.0000 - fn: 524.0000 - accuracy: 0.6257 - precision: 0.6257 - recall: 0.6257 - auc: 0.7663 - prc: 0.7940 - val_loss: 0.6526 - val_cross entropy: 0.6526 - val_tp: 86.0000 - val_fp: 70.0000 - val_tn: 86.0000 - val_fn: 70.0000 - val_accuracy: 0.5513 - val_precision: 0.5513 - val_recall: 0.5513 - val_auc: 0.6744 - val_prc: 0.7179\n",
            "Epoch 3/10\n",
            "44/44 [==============================] - 2s 48ms/step - loss: 5.8266e-04 - cross entropy: 0.4847 - tp: 1029.0000 - fp: 371.0000 - tn: 1029.0000 - fn: 371.0000 - accuracy: 0.7350 - precision: 0.7350 - recall: 0.7350 - auc: 0.8354 - prc: 0.8432 - val_loss: 0.4093 - val_cross entropy: 0.4093 - val_tp: 136.0000 - val_fp: 20.0000 - val_tn: 136.0000 - val_fn: 20.0000 - val_accuracy: 0.8718 - val_precision: 0.8718 - val_recall: 0.8718 - val_auc: 0.9481 - val_prc: 0.9491\n",
            "Epoch 4/10\n",
            "44/44 [==============================] - 2s 46ms/step - loss: 4.3355e-04 - cross entropy: 0.3931 - tp: 1147.0000 - fp: 253.0000 - tn: 1147.0000 - fn: 253.0000 - accuracy: 0.8193 - precision: 0.8193 - recall: 0.8193 - auc: 0.9049 - prc: 0.9057 - val_loss: 0.3507 - val_cross entropy: 0.3507 - val_tp: 141.0000 - val_fp: 15.0000 - val_tn: 141.0000 - val_fn: 15.0000 - val_accuracy: 0.9038 - val_precision: 0.9038 - val_recall: 0.9038 - val_auc: 0.9515 - val_prc: 0.9441\n",
            "Epoch 5/10\n",
            "44/44 [==============================] - 2s 48ms/step - loss: 3.7358e-04 - cross entropy: 0.3095 - tp: 1247.0000 - fp: 153.0000 - tn: 1247.0000 - fn: 153.0000 - accuracy: 0.8907 - precision: 0.8907 - recall: 0.8907 - auc: 0.9471 - prc: 0.9408 - val_loss: 0.2474 - val_cross entropy: 0.2474 - val_tp: 148.0000 - val_fp: 8.0000 - val_tn: 148.0000 - val_fn: 8.0000 - val_accuracy: 0.9487 - val_precision: 0.9487 - val_recall: 0.9487 - val_auc: 0.9850 - val_prc: 0.9849\n",
            "Epoch 6/10\n",
            "44/44 [==============================] - 2s 55ms/step - loss: 3.0187e-04 - cross entropy: 0.2406 - tp: 1295.0000 - fp: 105.0000 - tn: 1295.0000 - fn: 105.0000 - accuracy: 0.9250 - precision: 0.9250 - recall: 0.9250 - auc: 0.9666 - prc: 0.9624 - val_loss: 0.4570 - val_cross entropy: 0.4570 - val_tp: 127.0000 - val_fp: 29.0000 - val_tn: 127.0000 - val_fn: 29.0000 - val_accuracy: 0.8141 - val_precision: 0.8141 - val_recall: 0.8141 - val_auc: 0.8798 - val_prc: 0.8684\n",
            "Epoch 7/10\n",
            "44/44 [==============================] - 2s 47ms/step - loss: 3.1862e-04 - cross entropy: 0.2615 - tp: 1269.0000 - fp: 131.0000 - tn: 1269.0000 - fn: 131.0000 - accuracy: 0.9064 - precision: 0.9064 - recall: 0.9064 - auc: 0.9548 - prc: 0.9476 - val_loss: 0.4471 - val_cross entropy: 0.4471 - val_tp: 132.0000 - val_fp: 24.0000 - val_tn: 132.0000 - val_fn: 24.0000 - val_accuracy: 0.8462 - val_precision: 0.8462 - val_recall: 0.8462 - val_auc: 0.8668 - val_prc: 0.8544\n",
            "Epoch 8/10\n",
            "44/44 [==============================] - 2s 46ms/step - loss: 3.0368e-04 - cross entropy: 0.2547 - tp: 1279.0000 - fp: 121.0000 - tn: 1279.0000 - fn: 121.0000 - accuracy: 0.9136 - precision: 0.9136 - recall: 0.9136 - auc: 0.9573 - prc: 0.9488 - val_loss: 0.2225 - val_cross entropy: 0.2225 - val_tp: 144.0000 - val_fp: 12.0000 - val_tn: 144.0000 - val_fn: 12.0000 - val_accuracy: 0.9231 - val_precision: 0.9231 - val_recall: 0.9231 - val_auc: 0.9798 - val_prc: 0.9801\n",
            "Epoch 9/10\n",
            "44/44 [==============================] - 2s 46ms/step - loss: 2.2510e-04 - cross entropy: 0.1986 - tp: 1311.0000 - fp: 89.0000 - tn: 1311.0000 - fn: 89.0000 - accuracy: 0.9364 - precision: 0.9364 - recall: 0.9364 - auc: 0.9769 - prc: 0.9723 - val_loss: 0.1948 - val_cross entropy: 0.1948 - val_tp: 143.0000 - val_fp: 13.0000 - val_tn: 143.0000 - val_fn: 13.0000 - val_accuracy: 0.9167 - val_precision: 0.9167 - val_recall: 0.9167 - val_auc: 0.9857 - val_prc: 0.9863\n",
            "Epoch 10/10\n",
            "44/44 [==============================] - 2s 45ms/step - loss: 1.9191e-04 - cross entropy: 0.1598 - tp: 1327.0000 - fp: 73.0000 - tn: 1327.0000 - fn: 73.0000 - accuracy: 0.9479 - precision: 0.9479 - recall: 0.9479 - auc: 0.9821 - prc: 0.9796 - val_loss: 0.1408 - val_cross entropy: 0.1408 - val_tp: 147.0000 - val_fp: 9.0000 - val_tn: 147.0000 - val_fn: 9.0000 - val_accuracy: 0.9423 - val_precision: 0.9423 - val_recall: 0.9423 - val_auc: 0.9920 - val_prc: 0.9923\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history[\"accuracy\"])\n",
        "plt.plot(history.history[\"val_accuracy\"])\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "H1FKdrnY829W",
        "outputId": "5b68bf3b-4b48-49dc-9c37-de47b0fa8255"
      },
      "execution_count": 53,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiMAAAGdCAYAAADAAnMpAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABTeklEQVR4nO3dd3hUZfrG8e9MeodUIISOdBJaQrGDYsMuWEFULIu7Kru6YEPXgmWXZX/KiiAoKxYs2BXUWCkSivQuJaGkEUglbeb8/jgkEKUkMMmZydyf68qVkzPtCYHMzXve93lthmEYiIiIiFjEbnUBIiIi4t0URkRERMRSCiMiIiJiKYURERERsZTCiIiIiFhKYUREREQspTAiIiIillIYEREREUv5Wl1AbTidTvbu3UtYWBg2m83qckRERKQWDMOgsLCQFi1aYLcff/zDI8LI3r17SUhIsLoMEREROQUZGRm0bNnyuLd7RBgJCwsDzG8mPDzc4mpERESkNgoKCkhISKh+Hz8ejwgjVZdmwsPDFUZEREQ8zMmmWGgCq4iIiFhKYUREREQspTAiIiIillIYEREREUspjIiIiIilFEZERETEUgojIiIiYimFEREREbGUwoiIiIhYSmFERERELKUwIiIiIpZSGBERERFLKYyIyLEVZsLCf0PuVqsrEZF6VlBagcNpWPb6HrFrr4g0IEcFLH0VfngOygthxWwYmwa+/lZXJiKnocLhJCOvhO05xWzPLWJHbjG/5RSzPaeY3KIyvvvrObSLCbWkNoURETli+w/w5UOQu/nIuQM7YNkMGDDWsrJEpHYMwyC3qJztOWbY2J5bzPacIrbnFJOeV0LlCUY/0vNKFEZExEL5u2HBI7DhY/Pr4GgY8gQ4K+Hz++HHFyDxBgiOtLBIEalyqNzBjtxiM3DkFJmh4/BxYWnlcR8X5OdD2+gQ2sWE0C46hHYxobSLCaFtdAhhgX4N+B3UpDAi4s0qy2DxS/Dzv6CiBGx26DcGzpsAQU3B6YC0GZC93gwkFz9ndcUiXsPpNNibf8i8rHI4cJjho5g9Bw8d93E2G7RsGkTb6FDaRYfQPsYMHW2jQ2gWHojdbqv5gP2/wap3IOVu88EWUBgR8VZbvob5f4e87ebXrQbCJS9Asx5H7mP3gaFPw5tXmZdq+t0B0R2sqVekkco/VFF9KcW8tHLkuKzSedzHRQT5HR7hCK0x0tE6KphAP58Tv+iBnbD+I1g3DzLXmOda9YcWvVz3jdWBwoiIt8nbAfMnwJavzK9Dm8GFT0OPa4/9v6L250PHC2Hr1/DtRLj+rYatV6QRqHA42bW/5MhllaMmkeYWlR/3cX4+NlpHmUGjbUwI7auCR0woTYP9sNVlJCN/N6z/GNbPgz0rjpy3+UC7c8B5/OBT3xRGRLxFeQksmgILp4CjDOy+0P8eOOfvEBB24sde8BRsS4VNn8POhdDmzIaoWMSjGIZBTlHZ4csqxTUmkabnlZxw6WxsWEB1yGhXPacjlJZNg/D1OY0uHIWZsOETcwQk45cj5212899xt6uhyzAIiT7113ABhRGRxs4wzBAx/2HITzfPtTsXLn4BYjrV7jliO0OfW2H5TFjwMIz5AexqUyTeq7TCwZLt+1m7O//IfI6cYgrLjj95NNi/avKoOX+j/eHA0TYmhNAAF74dF+XAxk/MUZCdC4GqEGSDVgOg+9XQ5XIIi3Pda54mhRGRxix3K3z1EPz2nfl1eEu46FnzF1FdJ6qdOwHWvAf7VsOauZB0g+vrFXFjuUVlfLcpm283ZLFwWy4l5Y4/3Kdq8mi7oy6ntD98iaVZeGDdLqvURUkebPzMvASz42cwjqqtZT9zBKTblRDeon5e/zQpjIg0RmVF8NOLsGQqOCvAxx8G3QdnPgD+Iaf2nKExcPZf4dsnIPUf0PUK8A92adki7sQwDLZmF/HNhixSN2bxa8ZBjKOutMSFBzCofTTtY0OrV6y0iqzF5FFXKc2HTV+Yl2C2f28uxa/SPMkcAel2FTRp1TD1nAaFEZHGxDBg3Yfw9WNQuNc813EoXDQJotqf/vOn3APLZpmXe5a8DOc8dPrPKeJGKhxO0nbk8e3GLL7dmEVGXs0ltN3jwxncOY4LusbRrUV4/Y10HE9ZIWyeb46AbPsWHEdNfo3rboaPble55t97A1IYEWkssjbAlw/CroXm103bwEXPQ6eLXPcafoEwZCJ8eLs5Ebb3SAhr5rrnF7FAfkkFP2zJ5psNWfy4JadG0zB/XzsD20cxpEscg7vE0jwiqOELLC+BrQvMEZCtX0Nl6ZHbojsdHgG5GmLOaPjaXERhRMTTHTpo7iOTNt28TuwbBGf9FQb+2QwPrtb9Glg6DXYvg++ehitedv1riNSznbnF1aMfy3YeqLHSJSrEn/M7xzK4SxxndYwmxJWTS2urohS2fWMGkC3zzaaEVSLbHwkgsV0sa1TmSgojIp7K6YTV75i9P4pzzHNdLoehz9TvNWKbDS58BmZdCL/OgZS7ajZKE3FDDqfByvQDZgDZkMVvOcU1bj8jLpTBXeIY0iWOpIQm+Py+S2lDqCw3J5uvnwebvjQ3qqzSpJUZPrpfDc16NooAcjSFERFPtHeVeUlmd5r5dVRHs3tq+/Mb5vVbpZjXpdd/ZO5pM/KTRvfLUTxfUVklP2/J4ZuNWfywOYe84iPzK3ztNpLbRjLkcABpFWXRZGxHBez4EdZ9BJs+MyelVgmPPzwH5GqI792o/40pjIh4kpI8+O4pWP46YIB/qDmJNOUe8PVv2FqGPGHO5N/xo3kd+4yhDfv6tVDhcPLFmn18sGI3wf4+pLSLIqVtJF2ah1vzP1+pd3sOHiJ1Yxbfbszml9/2U+440lU0PNCX8w5ffjnnjBgigizaGM7pMPt/rJ8HGz6FQ3lHbguNg65XmiMgLZO9pp+PwoiIJ3A6YOVsc0ntoQPmuR7XwQX/sK5vQNM25sZai/8Pvn7UHJXxsW7Xz6MVllYwd1kGsxbuYG/+kcl+X2/IAiAs0JfkNpGktIskpW0U3VqEn16XS6k9w4CNn0LmWgiJgeAo83PVR3CkuSdSLTmdBmv35JO6MYtvNmazcV9BjdvbRAUfnnwaR982TfGz6ufsdJodUNfNMzuiFmcfuS042lwq3/1qsylZHb7/xsJmGMbx+9O6iYKCAiIiIsjPzyc8PNzqckQaVkYafPk3s9kYQGw3uORFaDPI2rrAnDz7Um8o2Q+X/BOSx1haTmZ+Ka8v2sHbS9OrO2FGhwYwckBr/HzsLN2xn+U7D1D0uy6ZoQG+9GndtDqc9GwZYd2bVmP3w3Pww6QT3MFmBpKQGPNNOiT6qLBiBpeygChW5vrwzS4Hn285RPZRe7vYbdCnddPq+R/tY0IafvltFcMwJ3qvmwcbPobCfUduC2pqtmHvdjW0OQt8GufYQG3fvxVGRNxVUbbZYGzV4Y3pAiLg/Eeg7+3u9YsrbYYZloKj4C+/QmBEg5ewcV8BM37ezqer9lJ5eFVE+5gQ7jy7HVckxddoQlXpcLJhXwFLt+exdMd+0nbkUVBaM5wE+fmY4aRtJCntokhMiCDA1/v+t+pyP71orsACcy6EYUBxrjkBuyTXvAxJ3d6SKgwfDhBOWUAk/uGxNIlpQUBE3B9HXA4HGfxD63fuhWHA3l/NSzDrP4b8jCO3BURA50vNEZB257rNSGJ9UhgR8VSOSlj2Gnz/LJQdnszW62YY/ITZBdXdOCrglYGQu8Xs8nrBPxrkZQ3DYOG2XKb/tJ2ft+ZWn09pG8mdZ7fjvE6x2GsxL8ThNNiUWTOcHCipqHGfAF87vVo1IaVtFCntIundqmnDddlsLBZOMVd+AQx5Es68/4/3cVSa8yeKczGKs9m7J4Pfdu5k377dOAqyibIVEGUrIJICYuwFhFHyx+c4Gd/AY1we+t0ITPVt0eBXi74ihgFZ68wRkPUfwYEdR27zD4VOF5sjIB0Gg29A3Wv2YAojIp5o50L48iHIXm9+3TzJvPyR0M/Ssk5qywJ4e7jZdv7eZeZ8knpS4XDy+Zq9TP9pR/X8ALsNLunRnDFntSMxoclpPb/TabYAX7pjf3VA+f0W7/4+dhITIkhua17W6dO6qTW9KDzF4pfh60fM4/MfhbMfPObdyiudLN2xn283mBNQ9xys2f00sWVE9eWXLs3DsDnKzUuExTmHP3KPjLQcPeJS9XXFKYQX/7AjoypVASU4+sjx/m1mCNm/9chjfIPMZoPdroaOF9Qu0DRSCiMinqRgr9nCfd0H5tdBTWHwRLPDqSdMZjMM+N8V5sqablfBdW+4/CUKSyt4Ny2DWYt2sO/wpNQgPx9G9Evg9jPbkhBZP0szDcPgt5ziGuEkq6Csxn187Ta6x0eQ0i6S/m2j6NumKWGBjX8IvlZ+mQbz/24enzMezptQ4+YDxeV8vzmb1I3Z/Lglp8Z8ngBfO2d2iGZI1zjO7xxLXPhpNvErL64ZWEp+F1x+/9lZcfLnrOITYAaP7lfDGRed+h5QjYzCiIgnqCyHX/4LP74AFcWADfreZv7vMTjS6urqJnMtTDsLMOD2byAh2SVPu/fgId5YvJN3fjcpdfSgNtyU0oomwQ27pNkwDHbtLzkqnOT94X/wdht0axFRPeckuU0kEcFeGE6WvQZf/NU8PutvlJ/9MLnF5WQVlLJsZx7fbshm+a48jmp+SkxYAIMPL789s0M0Qf4WhXHDgLKC4weVklxzXldghNlssNPFEKj3p99TGBFxd9tS4au/HxnebZlsrpJpkWRpWaflk7FmV9aW/cxAchoTBTfsNSelfrb6yKTUDrGh3HlWO67o1cKtJpTuPlBSPWqydEceu/bXvBxgs0HnZuGktI2kf7tIkttGERnSwH1h6olhGOQfqiC7sIycoz7it8/lkp3PATDX/2qeq7ieA4cqj/kcnZuFcUFXc/ltz/iIWs31Ec+gMCLirg6mw4KHYeNn5tchseakz54jPL/BUcE+eKmPOcpz7SxzH5s6MAyDn7fmMuPnmpNS+7eL5K6z23POGTEe8Ua1L/8QaTvy+OVwQNn+u9bjYLYfr5oQm9I2ipgw95rYWFrhMINFkRkufh82cgpLq2+vcNR8G7nO5wee952B3WbwWuXFPF15M2D+3HztNmLCAugQG1q9+VzLphZ1P5V6pzAi4m4qSs0GYT9PhspDYPMx93U5d7wly2HrzY8vwPfPQEQrczJrLTbrK6+smpS6nU2Z5n4cPnbb4UmpbenZskk9F12/sgtLSduRVz16siWr6A/3aRcTQkrbKPofDifNIly/yaHTaZBXUl4zVBSVkV1QFTpKq4NHYemxRzGOJyLIj5iwAK72+Zm7D/wTOwZr469na+9HiQ0PIiYsgJiwAJoE+XlEoBTXUBgRcSebv4L54+HATvPrNmfBxS9AXFdLy6oX5SXm6EjhXrNl/JkPHPeuBaUVvLM0ndcX7SSzwJyUGuxvTkq9bVD9TUq12v6iMpbtrBo5yWNTZgG//03cOirYnHNyePTkRKMHJeWV1eHi9yMY2YWl1aMbuUXlNXanPRl/XzsxoQHVQSI27MhxTGgAseGBxIQFEB3qb142W/sBzBsDhtPsh3Ppvxr1fipycgojIu5g/28wfwJsXWB+HdYChj5tLvlrzL+kV70DH98NAeHw55V/6I+y9+AhXl+0g3fSMqpXT8SEHZ6Umtza6yZ75pdUkLYzj6XbzTkn6/fm8/vMEN8kiJR2kYQH+tUMGoVlFJc76vR6USH+NUJFTHjAMUJHIOGBvrXvXrr+I/jgdjAc5iqwy/7j+Zcd5bQpjIhYqbzYvByz+P/AUQ52Pxh4L5z1NwgItbq6+ud0woxzzRb2fW+HyyYDsG5PPq/9vJ3P1+yrnpTaMTaUMWe344ok95qUaqWC0gpW7DzA0h3mZZ21u/Or/7yOJ8jPh9hjhoqq0GGOYkSF+ru+1f3Gz+C9UWYQSboJLn9ZQUSAeg4jU6dO5cUXXyQzM5PExEReeuklkpOPvYyvoqKCSZMmMXv2bPbs2UOnTp14/vnnueiii1z+zYhYrrQAVr0Ni1+Cgt3mufaDzUsy0R2sra2h7fgZZl+GYfNh+SWfM2W1nUXb9lffPLB9FGPObse5Z8RYt3eIhyguq2Rl+gGW7TxApcN5OGgEHgkaYQGE+PtY8+e4+SuYe4vZk6Pn9XDlfz2jN440iNq+f9e5ZeDcuXMZN24c06ZNIyUlhSlTpjB06FA2b95MbGzsH+7/6KOPMmfOHGbMmEHnzp1ZsGABV111FYsXL6ZXr151fXkR95SzGdKmw+p3ofzw5MSIVnDRJHMvCi98sy1PGEROs8HEZ6ZS+OkEFlU8hI/dxqWHO6X2aNmIJu3Ws5AAX87qGMNZHd1sO4Ct38B7I80g0v0aBRE5ZXUeGUlJSaFfv368/PLLADidThISEvjzn//M+PHj/3D/Fi1a8MgjjzB27Njqc9dccw1BQUHMmTOnVq+pkRFxS04HbJkPS181O49Wiels7l6bdJNXtoHOP1TBO2npvL5oB8GFO/na/yH8bA7mdPwP514yXMs4G4ttqfDODeAog65XwDWz3GsDR3EL9TIyUl5ezooVK5gw4Ug7X7vdzpAhQ1iyZMkxH1NWVkZgYM0lakFBQSxcuPC4r1NWVkZZ2ZF2ywUFBXUpU6R+leTByv/BspmQn26es9mh0yWQfCe0PdsrR0J2Hyjh9UU7eTctvXpCZWxYGzbGDqfnnne4OX86RNxicZXiEtt/gHdvNINI58vgmpkKInJa6vS3Jzc3F4fDQVxcXI3zcXFxbNq06ZiPGTp0KJMnT+bss8+mffv2pKamMm/ePByO48/+njRpEk8++WRdShOpf/vWmJdi1r4PleYyVIKaQu9R0O92aNLK2vossm5PPjMOT0qtWjbaKS6MMWe34/LEFviX94b/+9Lc/O/XOdBnlMUVy2nZuRDevt78N3DGxXDt6+DjXaufxPXqPcr+5z//YcyYMXTu3BmbzUb79u0ZPXo0s2bNOu5jJkyYwLhx46q/LigoICEhob5LFfkjR4W5UiBtOqQfNfrXrKfZsKz7NV55KcYwDH7YksOMn7az+Lcjk1IHdYhizFntOOfoSam+kXDO382us98/Y24kFhBmUeVyWnYtgbeGm037OlwAw2eDb+Noay/WqlMYiY6OxsfHh6ysrBrns7KyaNas2TEfExMTw8cff0xpaSn79++nRYsWjB8/nnbt2h33dQICAggIcK/WyOJlirJhxRuwfBYU7jPP2X3Na+PJd5mbwHnhpZiySgefrtrLaz/vYHPWkU6pw3o2546z2tE9/jiTUvuNMTdNy9sOi/5jbgQoniUjDd661mz13+48GDEHfPV7WlyjTmHE39+fPn36kJqaypVXXgmYE1hTU1O59957T/jYwMBA4uPjqaio4MMPP2T48OGnXLRIvdm93BwFWf+R2R8EzL1j+t4GfW6F8OaWlmeV/EMVvLV0F28s2kl2oTmfK8TfhxuSWzH6zLbENznJ6JCvPwx5Et67xVz23OdWiGhZ/4WLa+xeAXOuMVeKtT0bbninVm3+RWqrzpdpxo0bx6hRo+jbty/JyclMmTKF4uJiRo8eDcDIkSOJj49n0qRJACxdupQ9e/aQlJTEnj17eOKJJ3A6nTz00EOu/U5ETlVlmRk+lr4Ke1ceOd+ynzkK0vUKrx2K3n2ghFkLdzJ32ZFJqXHhAYwe1JYbklsREVSHuQJdhkGrgZC+GFKfgqtfraeqxaX2roI5V0FZAbQeBDe865WXJqV+1TmMjBgxgpycHB5//HEyMzNJSkpi/vz51ZNa09PTsR/Vea+0tJRHH32U7du3ExoayiWXXMKbb75JkyZNXPZNiJyS/D3mZZgVb0DJ4R1ifQLMeSDJYyC+t6XlWe2DFbsZ/+Ga6s6fnZuFMeasdgxLbIG/7yl017TZYOgzMOM8WPOuOefGy/+M3d6+NfC/K6A0HxL6w43vgX+I1VVJI6R28OJdDAN2LYa0V2Hj52b7aoDweHNFTO9REBJtbY1uYPbinUz8dD0A/dtFcs+5HTi7Y7RrOnzOuxPWzDX/l33rF14598YjZK2HNy6DQ3nmKOEtH2nisdRZvXVgFfFI5SWw9j1ImwFZ646cb3OWOQrS6VL1SThs6vfbeHHBZgBuG9SWxy7r4to244Mfhw2fwK5FsOlz8/KNuJfsTTD7cjOItOgNN3+oICL1Sr99pXHL22Gu4vj1TXOoGcAvGHoONxuUxXWztj43YhgGLyzYzCs//AbAXwZ35IEhHV2/30lESxhwL/z8T/jmceg41Gvn5Lil3K0we5h56bJ5ItwyDwLVul/ql8KIND6GAb99Z46CbJkPHL4S2bSNucS0101mszKp5nQaTPx0PW/+sguAhy/pzJ1nt6+/FzzzfrOLbd52MywO+FP9vZbU3v7fzEszxdkQ1wNu+Vj/VqRBKIxI41FaYG5UlzYd9m89cr79YHOyZIch2sTrGCodTh76cA3zVu7BZoNnruzBjSn13E02IAzOfwQ+uw9+fB4Sr4fgyPp9TTmxvB3miEhRJsR2hZGf6GciDUZhRDxf7lYzgKx6B8rNRlz4h5kjIP3ugOiO1tbnxsoqHdz3zirmr8/Ex25j8vBErkiKb5gX73ULLJ1uton/6UVzh2OxxoFdZhAp2APRnWDkpxASZXVV4kUURsQzOR2w9WuzN8j274+cjz7DnAuSeL0m3J3EoXIHd81ZwU9bcvD3sfPyjb24sNuxOynXC7sPDH0a3rzKDJN9b4foDg33+mI6mGEGkfwMiOoIoz6D0BirqxIvozAinuXQAXOztbQZcHDX4ZM26HSxGULanaulorVQUFrB7W8sY9nOAwT5+TBjZF/O7GjBkub255t7nGz7Br6dCNe/1fA1eLOCvWYQObgLItuZQSQs7uSPE3ExhRHxDJnrzP89r3nP3KQLILAJ9B5p9gdp2sbK6jxKXnE5o2alsXZPPmGBvrwxuh99Wls4N+DCp80Jx5s+N3eEbXOmdbV4k8JMc7LqgR3QpLUZRLx0uwOxnsKIuC9HpfkGlTbd7ElRJa67OQrS4zrwD7auPg+UXVDKTa8tZWt2EZEh/vzvtuTjb27XUGI7Q59RZjfcBQ/DmB/AfgodXqX2irLNEZG83yCiFdz6ufYKEkspjIj7KcqBlW/A8tfNCXUANh+zOVbKXdBqgC7FnIKMvBJunrmUXftLaBYeyJw7UugQG2p1WaZzH4Y178O+1WZzusTrra6o8SrONRua5W6B8JZw62fQpJ5XT4mchMKIuI/CTPj2CVj34VE75saYO7z2vQ3CW1hZnUfbll3Eza8tJbOglFaRwbx1RwoJkW40qhQaA2eNg9Qn4dsnocvlGvWqDyV55l4zORshrDmM+lSXOMUtKIyI+/jxeVj9jnkc38fcMbfbleAbYGlZnm793nxGzkxjf3E5HWNDmXNHCnHhbrj9e/8/maNh+emwZCqc86DVFTUuhw6YQSRrHYTGwajPIaoeG9uJ1IEuzIr72LXY/Hz5SzDmO0gcoSBymlbsOsD1039hf3E53ePDmXvXAPcMIgB+gTBkonm88N/mSJm4xqGD5hLqzDXmaOOoz7WMWtyKwoi4h5I8yNlkHne61NpaGolF23K5ZeZSCksr6demKW+P6U9kiJvvAdP9GojvCxXF8N3TVlfTOJQWwJxrYO+vEBxlrpqJOcPqqkRqUBgR95D+i/k5upM6P7rANxuyGP36MkrKHZzVMZr/3ZZCeKCf1WWdnM0GQ581j3+dA5lrra3H05UVwlvXwp7l5h4zIz+B2C5WVyXyBwoj4h7Sl5ifW/W3to5G4JNVe7h7zgrKHU6GdovjtVF9CfL3oD15WqVA1ysBA75+1Nz4UOquvBjeGg4ZS81dd0d+As16WF2VyDEpjIh7qA4jA6ytw8O9vTSd++euwuE0uLpXPFNv7E2ArwcFkSpDngAff9j+A2z9xupqPE95Cbw9AtIXQ0CEuftu80SrqxI5LoURsV55CexdZR63Vhg5VTN+2s7DH63FMOCW/q3553WJ+Pp46D/xyLaQcrd5/PUj4Kiwth5PUnEI3r0Bdv5sbhh5yzyI7211VSIn5KG/qaRR2bsSnBVm34Mmra2uxuMYhsG/v9nCM19uBODuc9rzjyu6Ybd7eGO4s/4KQZFmc64Vb1hdjWeoKIW5N5sjSv6hcPOH0LKv1VWJnJTCiFhv11HzRdRZtU4Mw+DpLzbyn9StADw4tBPjL+6MrTH8OQY1gfMeNo9/mASl+ZaW4/Yqy+C9kbDtW/ALhpveN+ffiHgAhRGxXvV8kYHW1uFhHE6DCfPWMnPhDgCeGNaVsec1st4RfW41t7Uv2Q8//8vqatyXowLeHw1bF4BvENw4F1rr35N4DoURsZbTARlp5rFW0tRahcPJfe/+yrvLMrDb4MVre3LroLZWl+V6Pn7mrr4Av7wCB3ZaWo5bclTAB7fB5i/ANxBueAfanm11VSJ1ojAi1spaB+WFEBAOcd2srsYjlFY4uPvNFXy+Zh9+PjZevrE31/VNsLqs+nPGUGh7jrlf0bdPWl2Ne3FUwrw7YeOn5uqj69+C9udZXZVInSmMiLWqmp0lJIPdA5egNrDiskpue2MZqZuyCfC1M31kXy7p0dzqsuqXzQZDnwFssH7ekZE0b+d0wMf3mH8mdj8YMQc6DLG6KpFTojAi1lKzs1rLL6ng5plLWfzbfkIDfJl9WzLndYq1uqyG0awH9LrJPF7wsBqhOZ3wyb2w9j2w+8Lw2eYIkoiHUhgR6xjGUStp1F/kRHIKyxgxfQm/ph+kSbAfb92RQv92XtY2/7xHzVUiu5eZowHeyumEz/4Cq98Gmw9cOws6az8n8WwKI2KdAzuhKNMcYo7vY3U1bmvvwUOMeHUJmzILiQ4NYO6dA0hMaGJ1WQ0vvDkMut88/vYJs6eGtzEM+GIc/Pom2OxwzQzoeoXVVYmcNoURsU7VfJEWvcAvyNpa3NTO3GKum7aE7bnFxDcJ4v27B9CpWZjVZVln4L0Q1gIOpsPSaVZX07AOHYDP74cVrwM2uOpVc5djkUZAYUSso/kiJ7Q5s5DrXl3CnoOHaBcdwvt3D6BtdIjVZVnLPwQGP2Ye//wvKM61tp6GkLUePrsPJnc93InWBlf+F3oOt7oyEZdRGBHrVIURNWf6g9UZBxkxfQk5hWV0bhbG3LsG0KKJRo8A6Hm9uelbWYHZmbUxclTChk/gjcvglYFmCKkogdhu5qqZpButrlDEpXytLkC8VHGuuecIQIJaVh9t6fb93D57OUVllSQlNGH26GQigv2sLst92O1w4TMw+zJY/jok3wkxnayuyjWKc83gsXwWFOwxz9l8oMtlkHyXGdwbQ6t/kd9RGBFrZCw1P8d0huBIa2txIz9szuauN1dQVulkQLsoZozqS2iA/pn+QduzoNOlZtfRrx+Dm96zuqLTs2clpM2AdR+Co8w8FxxttsPvOxoiWlpankh90285scauxeZnzRep9uXafdz37q9UOAwGd45l6k29CfRTI7jjuuAf5l4sWxfAb99B+/OtrqhuKsvNSzFpr5rLlau06A0pd0HXK8Ev0LLyRBqSwohYo2oljTbHA+CDFbt56IPVOA24rGdz/j0iCT8fTek6oegO0O8Oc1XNgkfh7p89o4tvwT5zRczy16E42zxn94PuV5uXnFr2tbY+EQsojEjDKy+BfavMY42MMHvxTiZ+uh6A6/sl8MxVPfCxa15ArZzzd1j9DmSvh1VvQe+RVld0bIZhXppc+qq5j4yz0jwf1hz63mZejgn1km66IsegMCINb89y85dxWAto0srqaiw19fttvLhgMwC3n9mWRy/tgk0TFGsvOBLOfgi+fgS+exq6XQ0BoVZXdUTFIVj7AaRNh8w1R863GmCOgnQZZu5MLOLlFEak4VVdomk9wGtXBhiGwQsLNvPKD78BcN/gjtw/pKOCyKlIHgPLXoMDO2DRf+D8R6yuyGzKtmwmrPwfHMozz/kGQo/rzBDSvKe19Ym4GYURaXjp3r0fjdNpMPHT9bz5yy4AHrmkC2PObmdxVR7MN8CczPreLbD4JegzyprVJ4YBO34yR0E2fwmG0zwf0QqS74Bet2jlmMhxKIxIw3JUHtkC3gvni1Q6nDz04RrmrdyDzQbPXNmDG1O8+1KVS3QZZk6GTl8MqU/B1a823GuXFcGad82luTmbjpxvd645CnLGRZ4xsVbEQgoj0rCy1kF5EQREQGxXq6tpUGWVDu57ZxXz12fiY7cxeXgiVyTFW11W42CzwdCnYcb5ZjBIuQvie9fva+7/zQwgq94yu8EC+IVA0g2NqxGbSANQGJGGVTVfJCHZq/63eKjcwV1zVvDTlhz8fexMvak3F3SNs7qsxiW+D/QYDmvfg68fhVu/cP2cJKcTtn1r9gbZ9u2R85HtzQCSdAMERrj2NUW8gMKINKx072t2VlBawe1vLGPZzgME+fkwY2RfzuwYbXVZjdPgx82ls7sWwaYvzDbqrnDoIKx6G5bNgLzth0/aoOOFkHIntDvfbFMvIqdEYUQajmEctZLGO5qd5RWXM2pWGmv35BMW6Msbo/vRp7UmMdabJgkw4F74+Z/wzWNmWPD1P/Xny95oTkhdPRcqis1zARHQ+xbodztEauKxiCsojEjDObADirLAx99sed3IZRWUcvNrS9maXURUiD//uz2Zbi00hF/vzrzfXFKbt91c8jvgT3V7vKMStnxlNijb+fOR87FdzWXEPUeAf4hLSxbxdgoj0nB2HV7S26JXo99zIyOvhJteW0p6XgnNwgOZc0cKHWLdqBlXYxYQZvYa+ew++PF5SLy+dktqi/fDytnmjrn5GeY5mx06X2rumNvmTK/tiyNS3xRGpOF4SX+RHbnF3DD9FzILSmkVGcxbd6SQEBlsdVnepdct5shG9gb46UW4aNLx77t3lbkqZu37R+2YGwW9R5mt2pskNEjJIt5MYUQaTvXmeI03jGQXlnLLzKVkFpTSMTaUOXekEBfeuEeB3JLdBy58GuZcbQaNfndAVPsjt1eWmxNd06abe8ZUaZ5kLgvudnWjH70TcScKI9IwinNh/1bzOCHZ2lrqSWFpBaNfX8buA4doHRXMO3f2Jzo0wOqyvFeHwdDhAtj2DXzzOFz/FhRmHdkxtyjTvJ/dD7pdeXjH3H66FCNiAYURaRhVl2hiujTKltjllU7umbOS9XsLiA7153+3JSuIuIMLn4LfUmHT5/DWcPjtO3BWmLeFNjuyY26Yer6IWElhRBrG0ZvjNTJOp8GDH6xm4bZcgv19eP3WZFpHabWFW4jtYoaN5bNg6wLzXEJ/c1VMl8tPb9mviLiMwog0jEY8efW5+Zv4ZNVefO02pt3chx4ttXzXrZz3KBTsg5Ao6DcGWiRZXZGI/I7CiNS/8mLYt9o8bmSdV1/7eTvTfzI7cr5wbU/OPiPG4orkD0Ki4MZ3ra5CRE5A/Yul/u1eDs5KCG8JTRrPDrWfrt7L019sBGD8xZ25urcF29aLiDQCCiNS/6qX9DaeUZHF23L563urALh1YBvuOlttwUVETpXCiNS/RrY53vq9+dz55goqHAaX9mjO45d1xabloCIip+yUwsjUqVNp06YNgYGBpKSkkJaWdsL7T5kyhU6dOhEUFERCQgIPPPAApaWlp1SweBhHJWQsM48bweZ4GXkl3Pr6MorKKunfLpJ/DU/EblcQERE5HXUOI3PnzmXcuHFMnDiRlStXkpiYyNChQ8nOzj7m/d9++23Gjx/PxIkT2bhxIzNnzmTu3Lk8/PDDp128eICsteZupwERZo8RD3aguJxRr6eRU1hG52ZhTB/Zl0A/H6vLEhHxeHUOI5MnT2bMmDGMHj2arl27Mm3aNIKDg5k1a9Yx77948WIGDRrEjTfeSJs2bbjwwgu54YYbTjqaIo1E1eZ4rVLA7rlXBQ+VO7ht9jK25xQT3ySIN0YnEx7oZ3VZIiKNQp3eHcrLy1mxYgVDhgw58gR2O0OGDGHJkiXHfMzAgQNZsWJFdfjYvn07X375JZdccslxX6esrIyCgoIaH+KhGkF/kUqHk3vfXsmv6QeJCPJj9m39aBahfUtERFylTn1GcnNzcTgcxMXVbJ0cFxfHpk2bjvmYG2+8kdzcXM4880wMw6CyspK77777hJdpJk2axJNPPlmX0sQdGYbHb45nGAaPfryO1E3ZBPjamXVrXzrEhlldlohIo1Lv4+Y//PADzz77LP/9739ZuXIl8+bN44svvuCpp5467mMmTJhAfn5+9UdGRkZ9lyn1IW87FGeDjz+06GV1Nafk399u5d1lGdht8NINvejTuvHtqyMiYrU6jYxER0fj4+NDVlZWjfNZWVk0a9bsmI957LHHuOWWW7jjjjsA6NGjB8XFxdx555088sgj2I8xjyAgIICAAG0y5vGqLtHE9/HI7djfWrqL/0s1dxp+6sruXNjt2H/HRUTk9NRpZMTf358+ffqQmppafc7pdJKamsqAAccehi8pKflD4PDxMVcgGIZR13rFk1TPF/G8/iJfr8/ksY/XAfCXwR25KaW1xRWJiDRedd6bZty4cYwaNYq+ffuSnJzMlClTKC4uZvTo0QCMHDmS+Ph4Jk2aBMCwYcOYPHkyvXr1IiUlhW3btvHYY48xbNiw6lAijdQuz5y8unxnHn9+51ecBlzfL4EHhnS0uiQRkUatzmFkxIgR5OTk8Pjjj5OZmUlSUhLz58+vntSanp5eYyTk0UcfxWaz8eijj7Jnzx5iYmIYNmwYzzzzjOu+C3E/RdmQ9xtgg4Rkq6uptW3Zhdw+ezlllU4Gd47l6Su7q7uqiEg9sxkecK2koKCAiIgI8vPzCQ8Pt7ocqY0Nn8J7t0BsN/jTYqurqZXM/FKueWUxew4eolerJrx9R3+C/DV6JyJyqmr7/u25XajEvXnY5nj5hyq49fU09hw8RLvoEGaO6qcgIiLSQBRGpH5Ub47n/vNFyiod3PXmcjZlFhITFsDs25KJDPG3uiwREa+hMCKuV1YE+9aYx63dO4w4nQbj5q7ml+15hAb48sbofiREBltdloiIV1EYEdfbsxwMB0QkQERLq6s5LsMw+MfnG/hi7T78fGxMv6UP3VpEWF2WiIjXURgR19vlGf1FXv1pO28s3gnAv4YnMbBDtLUFiYh4KYURcT0P2Bxv3srdPPeVuZ/So5d24fLEFhZXJCLivRRGxLUcFbB7uXnspmHkxy05PPSBOadlzFltueOsdhZXJCLi3RRGxLUy10BFMQQ2gZjOVlfzB2t353PPnBVUOg2uSGrBhIu7WF2SiIjXUxgR1zq6v8gxNkG00q79xYx+I42ScgdndojmxWsTsdvVXVVExGru9W4hns9NN8fLLSpj5Kw0covK6do8nFdu7o2/r/76i4i4A/02FtcxDLfcHK+4rJLb3ljGrv0lJEQG8cZt/QgL9LO6LBEROUxhRFxn/29Qkgs+AdCil9XVAFDhcHLPWytZszufyBB/Zo9OJjYs0OqyRETkKAoj4jpVl2ji+4BvgLW1YDY1+/uHa/hpSw5Bfj7MHNWXdjGhVpclIiK/ozAiruNm80VeXLCZeSv34GO3MfWmXvRq1dTqkkRE5BgURsR1qsJI64HW1gHMXryT//7wGwCTru7B+Z3jLK5IRESOR2FEXKMwC/K2AzZo2c/SUr5cu48nPlsPwF8vOIPhfRMsrUdERE5MYURco2pUJK4bBDWxrIxftu/n/ndXYRhwc/9W3Ht+B8tqERGR2lEYEdc4utmZRTZlFjDmf8spdzgZ2i2OJy/vjs2mpmYiIu5OYURcw+LN8fYePMSts5ZRWFpJ39ZN+c/1vfBRd1UREY+gMCKnr6zQ3JMGLAkjB0vKGTkrjcyCUjrGhvLaqL4E+vk0eB0iInJqFEbk9O1eBoYTIlpBRHyDvnRphYM7Zi9nW3YRzcIDmX1bMk2C/Ru0BhEROT0KI3L6quaLtG7YURGH0+Av7/zK8l0HCAv0ZfZtybRoEtSgNYiIyOlTGJHTZ0GzM8MwePyTdXy9IQt/XzszRvalU7OwBnt9ERFxHYUROT2OCti93DxuwPkiL3+3jbeWpmOzwX9GJNG/XVSDvbaIiLiWwoicnn1roKIEgppCdKcGecn3lmXwr2+2APDEsG5c3KN5g7yuiIjUD4UROT1Vl2gS+oO9/v86fbcpiwkfrQXgT+e2Z9TANvX+miIiUr8URuT0NOB8kV/TD/Cnt1bicBpc07slDw5tmJEYERGpXwojcuoMo8E2x9ueU8RtbyyjtMLJOWfE8Nw1PdRdVUSkkVAYkVO3fxuU7AffQGieWG8vk11QyshZaRwoqaBnywj+e1Nv/Hz0V1dEpLHQb3Q5dbsWm5/j+4BvQL28RGFpBbe+vozdBw7RJiqYWbf2IyTAt15eS0RErKEwIqeuenO8+lnSW17p5O45K9iwr4DoUH9m35ZMdGj9hB4REbGOwoicunrcHM/pNPjb+6tZtG0/If4+vH5rMq2jQlz+OiIiYj2FETk1hZlwYAdgg4R+Ln/6SV9t5NPVe/G123jl5j70aBnh8tcQERH3oDAip6ZqVCSuOwS6Nii89vN2Zvy8A4AXru3J2WfEuPT5RUTEvSiMyKmpp83xPlm1h6e/2AjA+Is7c3Xvli59fhERcT8KI3JqqlbSuLDZ2S/b9/O391cDcOvANtx1djuXPbeIiLgvhRGpu9ICyFpnHrto8mp+SQX3v7uKCofBpT2a8/hlXdXUTETESyiMSN3tXgaGE5q0hvAWLnnKxz9dR2ZBKW2jQ3jxup7Y7QoiIiLeQmFE6s7F/UU+X7OXT1btxW6DycMTCfZXUzMREW+iMCJ158LN8bILSnn0Y/OSz9jzOtCrVdPTfk4REfEsCiNSN5XlsHu5eXyam+MZhsFDH67hYEkF3ePD+cvgji4oUEREPI3CiNRN5hqoPARBkRB9xmk91VtL0/lhcw7+vnb+PTxJm9+JiHgp/faXujl6Se9prHbZkVvMM4f7ifz9os50jAtzRXUiIuKBFEakblwwebXS4WTce6s4VOFgYPsoRg9s45raRETEIymMSO0Zhks2x5v242/8mn6QsABfXrwuUct4RUS8nMKI1F7uFjiUB75B0DzxlJ5i3Z58pny7FYAnr+hGfJMgV1YoIiIeSGFEaq9qVKRlX/D1r/PDSyscPDB3FZVOg4u7N+OqXvEuLlBERDyRwojUXvV8kVPrL/LPBZvZml1EdGgAz1zVQ+3eRUQEUBiRujiNzfGW/LafmYt2APDCtT2IDKn7yIqIiDROCiNSOwV74eAusNmhZXLdHlpawd/eX41hwA3JCZzfOa6eihQREU+kMCK1U3WJJq47BIbX6aFPfrqBPQcP0SoymEcv7VoPxYmIiCdTGJHaOcUlvfPXZfLhyt3Vm+CFBGgTPBERqUlhRGrnFDbHyy4s5eGP1gJw1znt6dsmsj4qExERD6cwIidXmg9Z683jWo6MGIbBhA/XkldcTpfm4Tww5PT2sRERkcZLYUROLmMZGE5o2gbCm9fqIXOXZZC6KRt/HztTRiTh76u/aiIicmyn9A4xdepU2rRpQ2BgICkpKaSlpR33vueeey42m+0PH5deeukpFy0NrI7zRdL3l/DU5xsA+NvQM+jUTJvgiYjI8dU5jMydO5dx48YxceJEVq5cSWJiIkOHDiU7O/uY9583bx779u2r/li3bh0+Pj5cd911p128NJA6bI7ncBr89f1VFJc7SG4Tye1ntqvn4kRExNPVOYxMnjyZMWPGMHr0aLp27cq0adMIDg5m1qxZx7x/ZGQkzZo1q/745ptvCA4OVhjxFJXlsGe5eVyLMDLj5+0s23mAEH8f/jU8ER9tgiciIidRpzBSXl7OihUrGDJkyJEnsNsZMmQIS5YsqdVzzJw5k+uvv56QkJC6VSrW2LcKKkshOAqiO57wrhv3FTD56y0ATBzWjYTI4AYoUEREPF2dmj7k5ubicDiIi6vZQTMuLo5Nmzad9PFpaWmsW7eOmTNnnvB+ZWVllJWVVX9dUFBQlzLFlY6eL3KCvWTKKs1N8ModToZ0ieO6vi0bqEAREfF0DbrEYebMmfTo0YPk5BO3E580aRIRERHVHwkJCQ1UofxBLTfHm/zNFjZlFhIV4s9z12gTPBERqb06hZHo6Gh8fHzIysqqcT4rK4tmzZqd8LHFxcW8++673H777Sd9nQkTJpCfn1/9kZGRUZcyxVWczlqtpEnbkcf0n7YDMOnqHkSHBjREdSIi0kjUKYz4+/vTp08fUlNTq885nU5SU1MZMODEkxvff/99ysrKuPnmm0/6OgEBAYSHh9f4EAvkboFDB8A3CJonHvMuRWWV/PX9VRgGXNenJRd2O3EoFRER+b06bxQybtw4Ro0aRd++fUlOTmbKlCkUFxczevRoAEaOHEl8fDyTJk2q8biZM2dy5ZVXEhUV5ZrKpf5VjYq07As+fse8y1OfbSAj7xAtmwbx+DBtgiciInVX5zAyYsQIcnJyePzxx8nMzCQpKYn58+dXT2pNT0/Hbq854LJ582YWLlzI119/7ZqqpWGc5BLNtxuymLs8A5sN/nldImGBxw4sIiIiJ2IzDMOwuoiTKSgoICIigvz8fF2yaUhTesDBdLh5HnQYXOOm/UVlDJ3yE7lF5Yw5qy2PXKpRERERqam279/aMESOLX+PGURsdkioufrJMAwe/mgtuUXldIoL468XdrKoSBERaQwURuTYqi7RNOsBATX3lvlw5R4WrM/Cz8fG5BGJBPr5WFCgiIg0FgojcmzH2Y9m94ESnvh0PQD3DzmDbi0iGroyERFpZBRG5NiOEUacToO/vb+aorJK+rRuyt3ntLeoOBERaUwURuSPDh2ErHXm8VGdV2ct2sEv2/MI9vdhsjbBExERF1EYkT/avQwwoGlbCDObmG3JKuSFBZsBePTSrrSO0kaHIiLiGgoj8kdVk1dbDwSgvNLJ/e+uorzSyfmdY7khWXsFiYiI6yiMyB/tqmp2Zl6i+b/UrWzYV0DTYD9tgiciIi6nMCI1VZbBnhXmcasBrNh1gP/+sA2AZ67qQWxYoIXFiYhIY6QwIjXtXQWOMgiOpiSsDX99bxVOA67qFc8lPZpbXZ2IiDRCCiNSU/qRSzTPfLmJnftLaB4RyBOXd7O2LhERabQURqSmw2FkW1AP3lqaDpib4EUEaRM8ERGpHwojcoTTWd3s7Kk1TQAYPagNgzpEW1iUiIg0dgojckTuZig9SJktkEXFzekQG8rfL+psdVUiItLIKYzIEbsWA7C8sj3Y/fj38CRtgiciIvVOYUSqlfy2CIDlRif+MrgjPVpqEzwREal/CiMCmJvgFW9dCMD+qD786VxtgiciIg1DYUQA+OD7X4hxZFFp2Bk9/Fp8ffRXQ0REGobecYRt2UX88sMXAByM6Ezb+GYWVyQiIt5EYcTLVTicjHtvFUnGJgCiupxjcUUiIuJtFEa83MvfbWPN7nz6+24GwNZ6gMUViYiIt1EY8WKrMg7y8vfbCKeIjmSYJ1spjIiISMNSGPFSh8odjJu7CofT4J72edgwILI9hMZaXZqIiHgZhREv9dxXG9meW0xceACjW+41T2pURERELKAw4oV+3prD7CW7AHjx2kQC9y0zb2jV38KqRETEWymMeJn8kgoefH8NACMHtObstmGwZ4V5Y+uBFlYmIiLeSmHEyzz+6ToyC0ppGx3C+Is7w95fwVEOITEQ2c7q8kRExAspjHiRz9fs5ZNVe/Gx25g8PJFgf19IX2Le2Ko/2GzWFigiIl5JYcRLZBWU8shH6wAYe257erVqat6Q/ov5uZUu0YiIiDUURryAYRg8+MEa8g9V0CM+gj8P7mje4HRCRlUY0eRVERGxhsKIF5izNJ2ftuQQ4Gvn3yMS8avaBC9nI5Tmg18INOtpbZEiIuK1FEYauR25xTz7xUYA/n5RZzrEhh25sWq+SMu+4ONrQXUiIiIKI41apcPJA3NXcajCwaAOUdw6sE3NO1TNF9GSXhERsZDCSCM27cffWJVxkLBAX168NhG7/XerZXYdtZJGRETEIgojjdS6PflM+XYrAE9e3o0WTYJq3uFgBhTsBpsPxPe1oEIRERGTwkgjVFrh4IG5q6h0GlzcvRlX9Yr/452qLtE0T4SA0IYtUERE5CgKI43Qiws2szW7iOjQAJ65qge2YzUzS19sftbmeCIiYjGFkUZm8W+5zFy4A4AXru1BZIj/se+Yrv4iIiLiHhRGGpGC0gr+9t5qAG5IbsX5neOOfcdDByB7g3mskREREbGYwkgj8sSn69mbX0rrqGAevbTL8e+YvtT8HNUBQmMapjgREZHjUBhpJOav28e8lXuw22Dy8ERCAk7QxCxdS3pFRMR9KIw0AtmFpUyYtxaAu89pT5/WkSd+gDbHExERN6Iw4uFWZxzk+um/cKCkgi7Nw7l/yBknfkBFKexdaR5rZERERNyANiTxUBUOJy9/t42Xv9+Gw2kQFx7A/12fhL/vSfLl3pXgKIeQWIhs1zDFioiInIDCiAfall3EuPdWsWZ3PgDDElvw1BXdaBJ8nGW8R6uaL9J6AByr/4iIiEgDUxjxIE6nwewlO3nuq02UVTqJCPLjqSu7c3lii9o/SfV+NFrSKyIi7kFhxEPsPXiIBz9YzaJt+wE4q2M0L16bSLOIwNo/idMBGWnmseaLiIiIm1AYcXOGYfDxqj08/sl6CksrCfSz88glXbi5f+tjt3k/keyNUJYP/qEQ16N+ChYREakjhRE3dqC4nEc+XsuXazMBSEpowuThibSLOcWN7armi7TsBz760YuIiHvQO5Kb+n5TNg99uIacwjJ87TbuG9yRe85tj6/PaazGTtd8ERERcT8KI26muKySZ77cyNtL0wHoEBvKv4cn0aNlxOk9sWEcNXlV80VERMR9KIy4kRW78hj33mp27S8B4LZBbXnook4E+vmc/pPnZ0DhXrD7Qsu+p/98IiIiLqIw4gbKK51M+XYL0378DacBLSIC+ed1iQzsEO26F6kaFWmeCP4hrnteERGR06QwYrHNmYU8MHcVG/YVAHB173gmDutGRJCfa19I80VERMRNKYxYxOE0mLVwBy8u2Ey5w0nTYD+evaoHF/doXj8vWL05nsKIiIi4F4URC2TklfDX91eTtiMPgMGdY5l0TQ9iw+rQwKwuSvIgZ6N5rMmrIiLiZk5pnejUqVNp06YNgYGBpKSkkJaWdsL7Hzx4kLFjx9K8eXMCAgI444wz+PLLL0+pYE9mGAbvLc/g4v/8TNqOPIL9fXju6h68Nqpv/QURgIyl5ueojhDiwnkoIiIiLlDnkZG5c+cybtw4pk2bRkpKClOmTGHo0KFs3ryZ2NjYP9y/vLycCy64gNjYWD744APi4+PZtWsXTZo0cUX9HiO3qIwJ89byzYYsAPq2bsrk4Um0igqu/xc/enM8ERERN1PnMDJ58mTGjBnD6NGjAZg2bRpffPEFs2bNYvz48X+4/6xZs8jLy2Px4sX4+ZmTMtu0aXN6VXuYr9dnMmHeWvYXl+PnY2PcBZ248+x2+NgbaNdcbY4nIiJurE6XacrLy1mxYgVDhgw58gR2O0OGDGHJkiXHfMynn37KgAEDGDt2LHFxcXTv3p1nn30Wh8Nx3NcpKyujoKCgxocnKiyt4MH3V3PnmyvYX1xO52ZhfDL2TO45t33DBZGKQ7D3V/NY80VERMQN1WlkJDc3F4fDQVxcXI3zcXFxbNq06ZiP2b59O9999x033XQTX375Jdu2beNPf/oTFRUVTJw48ZiPmTRpEk8++WRdSnM7v2zfz1/fW82eg4ew2eDOs9sx7oIzCPB1QQOzutizEpwVENoMmrZt2NcWERGphXpfTeN0OomNjWX69On4+PjQp08f9uzZw4svvnjcMDJhwgTGjRtX/XVBQQEJCQn1XapLlFY4+NfXm3lt4Q4MAxIig/jXdUkkt420pqD0xebnVv2hrrv8ioiINIA6hZHo6Gh8fHzIysqqcT4rK4tmzZod8zHNmzfHz88PH58jIwJdunQhMzOT8vJy/P39//CYgIAAAgIC6lKaW1i/N59xc1ezOasQgBF9E3hsWFdCAyxcQa3+IiIi4ubqNGfE39+fPn36kJqaWn3O6XSSmprKgAHHfrMbNGgQ27Ztw+l0Vp/bsmULzZs3P2YQ8UQOp8HU77dx5dRFbM4qJDrUn9dG9uX5a3taG0ScDsg4vOxa80VERMRN1bnPyLhx45gxYwazZ89m48aN3HPPPRQXF1evrhk5ciQTJkyovv8999xDXl4e9913H1u2bOGLL77g2WefZezYsa77Liy0M7eY4a8u4cUFm6lwGAztFseC+89mSNe4kz+4vmWth7IC8A+DuO5WVyMiInJMdf5v+4gRI8jJyeHxxx8nMzOTpKQk5s+fXz2pNT09Hbv9SMZJSEhgwYIFPPDAA/Ts2ZP4+Hjuu+8+/v73v7vuu7CAYRi8nZbOM19spKTcQWiAL09c3o1resdjc5e5GVWXaBL6gY+a7YqIiHuyGYZhWF3EyRQUFBAREUF+fj7h4eFWl0N2QSkPfbiGHzbnANC/XST/vC6Rlk0boIFZXbw/GtbPg/MegXMesroaERHxMrV9/9Z/l+voizX7eOTjtRwsqcDf185DQztx26C22Buqb0htGYZ26hUREY+gMFJL+YcqmPjJOj5etReAbi3C+feIJM6IC7O4suM4uAsK94HdF+L7WF2NiIjIcSmM1MLCrbk8+MFq9uWXYrfBn87twF8Gd8Tf95T2GWwYVfNFmieBv5tdPhIRETmKwsgJHCp38Pz8TbyxeCcAbaKCmTwiid6tmlpbWG1oczwREfEQCiPHsTrjIOPeW8VvOcUA3Ny/FQ9f0oVgfw/5I9PmeCIi4iE85J214VQ4nEz9fhsvfbcNh9MgNiyAF67tybmdYq0urfaK90PuZvM4Qc3ORETEvSmMHGVbdhF/fW8Vq3fnA3Bpz+Y8fUV3moZ4WKfYjKXm5+hOEBJlbS0iIiInoTACOJ0G/1uyk0lfbaKs0kl4oC9PXdmdK5LirS7t1By9OZ6IiIib8/owsi//EA++v4aF23IBOKtjNC9c25PmEUEWV3YatDmeiIh4EK8NI4Zh8MmqvTz2yToKSysJ9LPz8CVduDmltfs1MKuL8hLYu8o81koaERHxAF4bRsoqnfzz680UllaSmNCEycMTaR8TanVZp2/PCnBWQFhzaNLa6mpEREROymvDSKCfD5OHJ7Hkt/2MPa89vj5u3MCsLqov0fQHd9mwT0RE5AS8NowAJLeNJLltpNVluJb2oxEREQ/TSIYDBACnAzLSzGOFERER8RAKI41J1jooLwT/MIjrZnU1IiIitaIw0phUzRdJSAa7j7W1iIiI1JLCSGOy63CzMy3pFRERD6Iw0lgYhpqdiYiIR1IYaSwO7ISiTLD7QXwfq6sRERGpNYWRxqJqVKRFL/Dz4Fb2IiLidRRGGgttjiciIh5KYaSx0HwRERHxUAojjUFxLuRuMY81MiIiIh5GYaQxqBoViekMwY2svb2IiDR6CiONwc6F5meNioiIiAdSGPF02Ztg+SzzuP1ga2sRERE5BQojnqyyHOaNAUcZdBgCXYZZXZGIiEidKYx4sh+fh8w1ENQULn8ZbDarKxIREakzhRFPlZEGCyebx5f9G8KbW1uPiIjIKVIY8URlRTDvTjCc0HMEdLvK6opEREROmcKIJ/r6UTiwA8Lj4eIXrK5GRETktCiMeJotC2DF6+bxlf+FoCaWliMiInK6FEY8SfF++ORe87j/n6DduZaWIyIi4goKI57CMODz+6A4G6I7weDHra5IRETEJRRGPMXqd2HjZ2D3haung1+Q1RWJiIi4hMKIJziYAV89ZB6fOx5aJFlajoiIiCspjLg7pxM+vgfKCqBlPxj0gNUViYiIuJTCiLv75b+w82fwC4arXgUfX6srEhERcSmFEXeWtQFS/2EeD30GotpbW4+IiEg9UBhxV5XlZpdVRxl0vBD6jLa6IhERkXqhMOKufpgEWWshKFKb4ImISKOmMOKO0n+BRVPM42FTICzOympERETqlcKIuykrhI/uMjfBS7wBul5hdUUiIiL1SmHE3Sx4BA7shIgEuPh5q6sRERGpdwoj7mTzV7ByNmCDK1+BwAirKxIREal3CiPuojgXPv2zeTxgLLQ9y9p6REREGojCiDswDPjsPijOgZgucP5jVlckIiLSYBRG3MGqt2HT52D3g6tfBb9AqysSERFpMAojVjuwC776u3l83gRonmhtPSIiIg1MYcRKToe5CV55ISSkwKD7ra5IRESkwSmMWGnJVNi1CPxC4KppYPexuiIREZEGpzBilaz18N1T5vFFz0JkO2vrERERsYjCiBUqyw5vglcOZ1wEvUdZXZGIiIhlFEas8P2zkLUOgqPg8pe0CZ6IiHg1hZGGtmsJLPqPeTzsPxAaa209IiIiFjulMDJ16lTatGlDYGAgKSkppKWlHfe+b7zxBjabrcZHYKCX9tGo2gQPA5Jugi7DrK5IRETEcnUOI3PnzmXcuHFMnDiRlStXkpiYyNChQ8nOzj7uY8LDw9m3b1/1x65du06raI81fwIc3AURreCi56yuRkRExC3UOYxMnjyZMWPGMHr0aLp27cq0adMIDg5m1qxZx32MzWajWbNm1R9xcXGnVbRH2vQF/PomYIOrXoHAcKsrEhERcQt1CiPl5eWsWLGCIUOGHHkCu50hQ4awZMmS4z6uqKiI1q1bk5CQwBVXXMH69etP+DplZWUUFBTU+PBoRTnw6V/M44H3Qpszra1HRETEjdQpjOTm5uJwOP4wshEXF0dmZuYxH9OpUydmzZrFJ598wpw5c3A6nQwcOJDdu3cf93UmTZpERERE9UdCQkJdynQvVZvgleRCbFc471GrKxIREXEr9b6aZsCAAYwcOZKkpCTOOecc5s2bR0xMDK+++upxHzNhwgTy8/OrPzIyMuq7zPrz6xzY/MXhTfCmaxM8ERGR3/Gty52jo6Px8fEhKyurxvmsrCyaNWtWq+fw8/OjV69ebNu27bj3CQgIICAgoC6luacDO2H+ePP4/EegWQ9LyxEREXFHdRoZ8ff3p0+fPqSmplafczqdpKamMmDAgFo9h8PhYO3atTRv3rxulXoapwM+ugfKi6DVABj4F6srEhERcUt1GhkBGDduHKNGjaJv374kJyczZcoUiouLGT16NAAjR44kPj6eSZMmAfCPf/yD/v3706FDBw4ePMiLL77Irl27uOOOO1z7nbibxS9B+mLwD9UmeCIiIidQ5zAyYsQIcnJyePzxx8nMzCQpKYn58+dXT2pNT0/Hbj8y4HLgwAHGjBlDZmYmTZs2pU+fPixevJiuXbu67rtwN5lr4bunzeOLJkHTNpaWIyIi4s5shmEYVhdxMgUFBURERJCfn094uJv356gsg+nnQfZ66HQJXP+29p4RERGvVNv3b+1N42rfPW0GkeBoGPZ/CiIiIiInoTDiSjsXmXNFAC7/PwiNsbYeERERD6Aw4iqlBfDR3YABvW6GzpdaXZGIiIhHUBhxlfkTID8dmrSCoZOsrkZERMRjKIy4wsbPYdUczE3wXtUmeCIiInWgMHK6irLhs8MNzQb9BVoPtLYeERERD6MwcjoMw9yNt2Q/xHWH8x6xuiIRERGPozByOlb+D7Z8BT7+5uUZ30awn46IiEgDUxg5VXk7YMHD5vH5j0Kz7tbWIyIi4qEURk6F02Eu4y0vgtaDYMC9VlckIiLisRRGTsWi/0DGL+AfBle+ok3wREREToPCSF3tWwPfP2seX/wcNG1tbT0iIiIeTmGkLipKYd6d4KyAzpdB0k1WVyQiIuLxFEbq4runIGcjhMTAsP9oEzwREREXUBiprR0/w5Kp5vHlL0FItLX1iIiINBIKI7VRmg8f3wMY0HskdLrY6opEREQaDYWR2vhqPORnQJPWMPRZq6sRERFpVBRGTmbDp7D6bbDZ4erpEBBmdUUiIiKNisLIiRRmwWf3mceD7oNW/a2tR0REpBFSGDkew4BP/wyH8iCuB5z7sNUViYiINEoKI8ez4g3YusDcBO/q6eDrb3VFIiIijZLCyLHkbYcFj5jHgx+HuK7W1iMiItKIKYz8nqMS5t0FFcXQ5izoP9bqikRERBo1hZHfWzQFdqdBQDhc+V+w649IRESkPumd9mj7VsMPk8zji1+AJq2srUdERMQLKIxUqd4ErxK6DIPE662uSERExCsojFRJ/QfkbIKQWLhMm+CJiIg0FIURgO0/wi+HN8G74mUIibK2HhERES+iMHLoIHz8J/O4z61wxlArqxEREfE6CiNf/R0KdkPTtnDhM1ZXIyIi4nW8O4ys/xjWvGtugnfVqxAQanVFIiIiXsd7w0h5CXz5N/P4zAegVYq19YiIiHgp7w0j/sEwYg50uRzOGW91NSIiIl7L1+oCLNWqv/khIiIilvHekRERERFxCwojIiIiYimFEREREbGUwoiIiIhYSmFERERELKUwIiIiIpZSGBERERFLKYyIiIiIpRRGRERExFIKIyIiImIphRERERGxlMKIiIiIWEphRERERCzlEbv2GoYBQEFBgcWViIiISG1VvW9XvY8fj0eEkcLCQgASEhIsrkRERETqqrCwkIiIiOPebjNOFlfcgNPpZO/evYSFhWGz2Vz2vAUFBSQkJJCRkUF4eLjLnldOjX4e7kc/E/ein4d70c/j5AzDoLCwkBYtWmC3H39miEeMjNjtdlq2bFlvzx8eHq6/SG5EPw/3o5+Je9HPw73o53FiJxoRqaIJrCIiImIphRERERGxlFeHkYCAACZOnEhAQIDVpQj6ebgj/Uzci34e7kU/D9fxiAmsIiIi0nh59ciIiIiIWE9hRERERCylMCIiIiKWUhgRERERS3l1GJk6dSpt2rQhMDCQlJQU0tLSrC7JK02aNIl+/foRFhZGbGwsV155JZs3b7a6LDnsueeew2azcf/991tditfas2cPN998M1FRUQQFBdGjRw+WL19udVley+Fw8Nhjj9G2bVuCgoJo3749Tz311En3X5Hj89owMnfuXMaNG8fEiRNZuXIliYmJDB06lOzsbKtL8zo//vgjY8eO5ZdffuGbb76hoqKCCy+8kOLiYqtL83rLli3j1VdfpWfPnlaX4rUOHDjAoEGD8PPz46uvvmLDhg3861//omnTplaX5rWef/55XnnlFV5++WU2btzI888/zwsvvMBLL71kdWkey2uX9qakpNCvXz9efvllwNz/JiEhgT//+c+MHz/e4uq8W05ODrGxsfz444+cffbZVpfjtYqKiujduzf//e9/efrpp0lKSmLKlClWl+V1xo8fz6JFi/j555+tLkUOu+yyy4iLi2PmzJnV56655hqCgoKYM2eOhZV5Lq8cGSkvL2fFihUMGTKk+pzdbmfIkCEsWbLEwsoEID8/H4DIyEiLK/FuY8eO5dJLL63x70Qa3qeffkrfvn257rrriI2NpVevXsyYMcPqsrzawIEDSU1NZcuWLQCsXr2ahQsXcvHFF1tcmefyiI3yXC03NxeHw0FcXFyN83FxcWzatMmiqgTMEar777+fQYMG0b17d6vL8VrvvvsuK1euZNmyZVaX4vW2b9/OK6+8wrhx43j44YdZtmwZf/nLX/D392fUqFFWl+eVxo8fT0FBAZ07d8bHxweHw8EzzzzDTTfdZHVpHssrw4i4r7Fjx7Ju3ToWLlxodSleKyMjg/vuu49vvvmGwMBAq8vxek6nk759+/Lss88C0KtXL9atW8e0adMURizy3nvv8dZbb/H222/TrVs3Vq1axf3330+LFi30MzlFXhlGoqOj8fHxISsrq8b5rKwsmjVrZlFVcu+99/L555/z008/0bJlS6vL8VorVqwgOzub3r17V59zOBz89NNPvPzyy5SVleHj42Nhhd6lefPmdO3atca5Ll268OGHH1pUkTz44IOMHz+e66+/HoAePXqwa9cuJk2apDByirxyzoi/vz99+vQhNTW1+pzT6SQ1NZUBAwZYWJl3MgyDe++9l48++ojvvvuOtm3bWl2SVxs8eDBr165l1apV1R99+/blpptuYtWqVQoiDWzQoEF/WOq+ZcsWWrdubVFFUlJSgt1e8+3Tx8cHp9NpUUWezytHRgDGjRvHqFGj6Nu3L8nJyUyZMoXi4mJGjx5tdWleZ+zYsbz99tt88sknhIWFkZmZCUBERARBQUEWV+d9wsLC/jBfJyQkhKioKM3jscADDzzAwIEDefbZZxk+fDhpaWlMnz6d6dOnW12a1xo2bBjPPPMMrVq1olu3bvz6669MnjyZ2267zerSPJfhxV566SWjVatWhr+/v5GcnGz88ssvVpfklYBjfrz++utWlyaHnXPOOcZ9991ndRle67PPPjO6d+9uBAQEGJ07dzamT59udUleraCgwLjvvvuMVq1aGYGBgUa7du2MRx55xCgrK7O6NI/ltX1GRERExD145ZwRERERcR8KIyIiImIphRERERGxlMKIiIiIWEphRERERCylMCIiIiKWUhgRERERSymMiIiIiKUURkRERMRSCiMiIiJiKYURERERsZTCiIiIiFjq/wFBZGpp/p0AaQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# normalization of our training data (note each feature individually processed!)\n",
        "test_feats_standardized = xr.concat([\n",
        "    z_score_testing(\n",
        "        train_feats.isel(feature=0).expand_dims(dim='feature', axis=3),\n",
        "        test_feats.isel(feature=0).expand_dims(dim='feature', axis=3)\n",
        "        ),\n",
        "    z_score_testing(\n",
        "        train_feats.isel(feature=1).expand_dims(dim='feature', axis=3),\n",
        "        test_feats.isel(feature=1).expand_dims(dim='feature', axis=3)\n",
        "        )],\n",
        "                                     dim='feature')\n",
        "\n",
        "test_label_ints = xr.where(test_label >= 75, 1, 0)\n",
        "test_y = onehot_layer(test_label_ints.astype(int))"
      ],
      "metadata": {
        "id": "zMbEUR_O83Aw"
      },
      "execution_count": 54,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate the model on the test data using `evaluate`\n",
        "print(\"Evaluate on test data\")\n",
        "results = model.evaluate(test_feats_standardized.values, test_y, batch_size=128)\n",
        "print(results)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KZ4Zljo_-a7c",
        "outputId": "023c8b41-3fa2-4654-8972-7bdedd526a3a"
      },
      "execution_count": 55,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluate on test data\n",
            "6/6 [==============================] - 0s 40ms/step - loss: 0.1566 - cross entropy: 0.1566 - tp: 634.0000 - fp: 34.0000 - tn: 634.0000 - fn: 34.0000 - accuracy: 0.9491 - precision: 0.9491 - recall: 0.9491 - auc: 0.9865 - prc: 0.9847\n",
            "[0.15658392012119293, 0.15658392012119293, 634.0, 34.0, 634.0, 34.0, 0.9491018056869507, 0.9491018056869507, 0.9491018056869507, 0.9865292310714722, 0.9847036004066467]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# generate predictions using the trained model\n",
        "predictions = model.predict(test_feats_standardized.values)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9fcwNVSZ-a_X",
        "outputId": "9490034c-af59-46aa-d963-986b0450af78"
      },
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "21/21 [==============================] - 0s 14ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred_ints = xr.where(predictions[:,1]>0.5, 1, 0)"
      ],
      "metadata": {
        "id": "tHLyxWLVAqed"
      },
      "execution_count": 57,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics.cluster import contingency_matrix\n",
        "contingency_matrix(test_label_ints.astype(int).values.squeeze(), pred_ints)\n",
        "# tend to predict more false positives"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JrZYzjM9_eLj",
        "outputId": "52c87e5b-750d-452e-97ba-8232aba8f32c"
      },
      "execution_count": 58,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[555,  21],\n",
              "       [ 13,  79]])"
            ]
          },
          "metadata": {},
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_label_ints.astype(int).sum().values"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UwAmxuqcBJ1K",
        "outputId": "73556e64-5a81-403c-a29f-9b3e349a163d"
      },
      "execution_count": 59,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(92)"
            ]
          },
          "metadata": {},
          "execution_count": 59
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred_ints.sum() - 92"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Bb-w77AwBfUV",
        "outputId": "ae962f47-9eaa-4f7e-e5ef-292abbab7b0b"
      },
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "8"
            ]
          },
          "metadata": {},
          "execution_count": 60
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Tutorial Practice: What happens if we do not use class weights during training?\n",
        "\n",
        "---"
      ],
      "metadata": {
        "id": "KmokO1cq9_RS"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# recreate the model object"
      ],
      "metadata": {
        "id": "0a43SpER-Mxd"
      },
      "execution_count": 61,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# recompile the model"
      ],
      "metadata": {
        "id": "ITEZzW3y-M02"
      },
      "execution_count": 62,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# retrain the model without class weights"
      ],
      "metadata": {
        "id": "Xak_WCCB-M3-"
      },
      "execution_count": 63,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# evaluate the new model's skill"
      ],
      "metadata": {
        "id": "qwdW6YEP-NCM"
      },
      "execution_count": 64,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## **Unsupervised Learning**\n",
        "\n",
        "---\n",
        "\n",
        "Unsupervised learning refers to machine learning tasks where the dataset does not contain predictands (y) and solely contains predictors (x). In other words, in unsupervised learning, we do not know the \"answers\" or have labels beforehand. Unsupervised learning is particularly useful for dimensionality reduction, anomaly detection, image denoising, and pattern grouping (e.g., clustering)."
      ],
      "metadata": {
        "id": "oOZ_y1G8dFgF"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![supervised](https://drive.google.com/uc?export=view&id=1HEcZ4VXU_Q9Y9rrgnRkAg4JdjRJSs9Nj)"
      ],
      "metadata": {
        "id": "2zqQVMG879C9"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Some examples of unsupervised learning problems in the environmental sciences include:\n",
        "\n",
        "*   uncovering common polynya event patterns from sea ice concentration in the Ross Sea, Antarctica (e.g., DuVivier et al. 2023)\n",
        "*   grouping high shear-low CAPE severe thunderstorm event environments (e.g., Campbell et al. 2024)\n",
        "*   identifying subseasonal weather regimes over North America (e.g., Molina et al. 2023)\n",
        "*   creating an index for a climate teleconnection, such as ENSO teleconnection to Southern California precipitation (e.g., Passarella and Mahajan 2023)\n",
        "\n"
      ],
      "metadata": {
        "id": "YEwMoWtbA7fV"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Unsupervised Learning Example\n",
        "\n",
        "---"
      ],
      "metadata": {
        "id": "FWOflQWICzK9"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "We will be training an autoencoder to perform the task of anomaly detection. For anomaly detection, we would note instances where reconstruction error for an image is higher, indicating that the sample was likely not similar to those in the training set distribution.\n",
        "\n",
        "We will again use the previous data, but this time we will focus on reconstructing one input variable (feature)."
      ],
      "metadata": {
        "id": "2I2BURFEDLj6"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 65,
      "metadata": {
        "id": "CF3Qy6sIOlqP"
      },
      "outputs": [],
      "source": [
        "features = xr.concat([\n",
        "    ds['ctt_grid'].expand_dims(dim='feature', axis=3)],\n",
        "          dim='feature')\n",
        "\n",
        "# note that we are reconstructing the respective input image/variable\n",
        "train_feats, _, test_feats, _ = random_split(features, features)\n",
        "\n",
        "# normalization of our training data (note each feature individually processed!)\n",
        "train_feats_standardized = xr.concat([\n",
        "    z_score(train_feats.isel(feature=0).expand_dims(dim='feature', axis=3))],\n",
        "                                     dim='feature')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "number_of_features = 1"
      ],
      "metadata": {
        "id": "yYD-l_b-E1jB"
      },
      "execution_count": 66,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = keras.Sequential()\n",
        "\n",
        "# input\n",
        "model.add(keras.Input(shape=(32, 32, number_of_features)))  # 32x32 images\n",
        "\n",
        "# encoder\n",
        "# hidden layers\n",
        "model.add(layers.Conv2D(16, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.MaxPooling2D(2))\n",
        "\n",
        "model.add(layers.Conv2D(8, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.MaxPooling2D(2))\n",
        "\n",
        "model.add(layers.Conv2D(4, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.MaxPooling2D(2))\n",
        "\n",
        "# decoder\n",
        "model.add(layers.Conv2D(4, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.UpSampling2D(2))\n",
        "\n",
        "model.add(layers.Conv2D(8, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.UpSampling2D(2))\n",
        "\n",
        "model.add(layers.Conv2D(16, 3, strides=1, padding=\"same\", activation=\"relu\"))\n",
        "model.add(layers.UpSampling2D(2))\n",
        "\n",
        "# Finally, we add last layer\n",
        "model.add(layers.Conv2D(1, (3, 3), activation='linear', padding='same'))\n",
        "\n",
        "# print model summary\n",
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "teWxwzffDhyu",
        "outputId": "a7e6fc2e-b8fd-4c1c-eaba-de04c2d7ea42"
      },
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_2\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d_6 (Conv2D)           (None, 32, 32, 16)        160       \n",
            "                                                                 \n",
            " max_pooling2d_6 (MaxPoolin  (None, 16, 16, 16)        0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_7 (Conv2D)           (None, 16, 16, 8)         1160      \n",
            "                                                                 \n",
            " max_pooling2d_7 (MaxPoolin  (None, 8, 8, 8)           0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_8 (Conv2D)           (None, 8, 8, 4)           292       \n",
            "                                                                 \n",
            " max_pooling2d_8 (MaxPoolin  (None, 4, 4, 4)           0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_9 (Conv2D)           (None, 4, 4, 4)           148       \n",
            "                                                                 \n",
            " up_sampling2d (UpSampling2  (None, 8, 8, 4)           0         \n",
            " D)                                                              \n",
            "                                                                 \n",
            " conv2d_10 (Conv2D)          (None, 8, 8, 8)           296       \n",
            "                                                                 \n",
            " up_sampling2d_1 (UpSamplin  (None, 16, 16, 8)         0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_11 (Conv2D)          (None, 16, 16, 16)        1168      \n",
            "                                                                 \n",
            " up_sampling2d_2 (UpSamplin  (None, 32, 32, 16)        0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_12 (Conv2D)          (None, 32, 32, 1)         145       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 3369 (13.16 KB)\n",
            "Trainable params: 3369 (13.16 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "METRICS = [\n",
        "    keras.metrics.MeanSquaredError(name='mean_squared_error'),\n",
        "    keras.metrics.RootMeanSquaredError(name='root_mean_squared_error'),\n",
        "]\n",
        "\n",
        "model.compile(\n",
        "    optimizer=keras.optimizers.Adam(learning_rate=0.001),\n",
        "    loss=\"mean_squared_error\",\n",
        "    metrics=METRICS\n",
        ")\n",
        "\n",
        "BATCH_SIZE = 32\n",
        "EPOCHS = 10"
      ],
      "metadata": {
        "id": "iutgSt7QDh77"
      },
      "execution_count": 68,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "history = model.fit(\n",
        "    x=train_feats_standardized.values,\n",
        "    y=train_feats_standardized.values,\n",
        "    batch_size=BATCH_SIZE,\n",
        "    epochs=EPOCHS,\n",
        "    validation_split=0.1,\n",
        "    shuffle=True,\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NJ2SPdP3HL6l",
        "outputId": "7962aefa-fe38-4057-f6fb-16faf90d9c28"
      },
      "execution_count": 69,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "44/44 [==============================] - 3s 40ms/step - loss: 0.6365 - mean_squared_error: 0.6365 - root_mean_squared_error: 0.7978 - val_loss: 0.4575 - val_mean_squared_error: 0.4575 - val_root_mean_squared_error: 0.6764\n",
            "Epoch 2/10\n",
            "44/44 [==============================] - 1s 27ms/step - loss: 0.1989 - mean_squared_error: 0.1989 - root_mean_squared_error: 0.4460 - val_loss: 0.0809 - val_mean_squared_error: 0.0809 - val_root_mean_squared_error: 0.2845\n",
            "Epoch 3/10\n",
            "44/44 [==============================] - 1s 27ms/step - loss: 0.0675 - mean_squared_error: 0.0675 - root_mean_squared_error: 0.2599 - val_loss: 0.0644 - val_mean_squared_error: 0.0644 - val_root_mean_squared_error: 0.2538\n",
            "Epoch 4/10\n",
            "44/44 [==============================] - 1s 26ms/step - loss: 0.0543 - mean_squared_error: 0.0543 - root_mean_squared_error: 0.2330 - val_loss: 0.0549 - val_mean_squared_error: 0.0549 - val_root_mean_squared_error: 0.2342\n",
            "Epoch 5/10\n",
            "44/44 [==============================] - 1s 27ms/step - loss: 0.0478 - mean_squared_error: 0.0478 - root_mean_squared_error: 0.2186 - val_loss: 0.0505 - val_mean_squared_error: 0.0505 - val_root_mean_squared_error: 0.2246\n",
            "Epoch 6/10\n",
            "44/44 [==============================] - 1s 26ms/step - loss: 0.0438 - mean_squared_error: 0.0438 - root_mean_squared_error: 0.2094 - val_loss: 0.0481 - val_mean_squared_error: 0.0481 - val_root_mean_squared_error: 0.2193\n",
            "Epoch 7/10\n",
            "44/44 [==============================] - 1s 27ms/step - loss: 0.0411 - mean_squared_error: 0.0411 - root_mean_squared_error: 0.2028 - val_loss: 0.0441 - val_mean_squared_error: 0.0441 - val_root_mean_squared_error: 0.2100\n",
            "Epoch 8/10\n",
            "44/44 [==============================] - 1s 27ms/step - loss: 0.0380 - mean_squared_error: 0.0380 - root_mean_squared_error: 0.1949 - val_loss: 0.0461 - val_mean_squared_error: 0.0461 - val_root_mean_squared_error: 0.2147\n",
            "Epoch 9/10\n",
            "44/44 [==============================] - 1s 26ms/step - loss: 0.0374 - mean_squared_error: 0.0374 - root_mean_squared_error: 0.1934 - val_loss: 0.0392 - val_mean_squared_error: 0.0392 - val_root_mean_squared_error: 0.1979\n",
            "Epoch 10/10\n",
            "44/44 [==============================] - 2s 35ms/step - loss: 0.0347 - mean_squared_error: 0.0347 - root_mean_squared_error: 0.1863 - val_loss: 0.0380 - val_mean_squared_error: 0.0380 - val_root_mean_squared_error: 0.1951\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history[\"mean_squared_error\"], c='k')\n",
        "plt.plot(history.history[\"val_mean_squared_error\"], c='r')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "m7Ysy6IFId2i",
        "outputId": "44d37a60-96d5-4074-894c-f73c4d159692"
      },
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# normalization of our training data (note each feature individually processed!)\n",
        "test_feats_standardized = xr.concat([\n",
        "    z_score_testing(\n",
        "        train_feats.isel(feature=0).expand_dims(dim='feature', axis=3),\n",
        "        test_feats.isel(feature=0).expand_dims(dim='feature', axis=3)\n",
        "        )],\n",
        "                                     dim='feature')"
      ],
      "metadata": {
        "id": "I3QHjf_UHOHC"
      },
      "execution_count": 71,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate the model on the test data using `evaluate`\n",
        "print(\"Evaluate on test data\")\n",
        "results = model.evaluate(test_feats_standardized.values, test_feats_standardized.values, batch_size=128)\n",
        "print(results)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UJknPtMCIRcg",
        "outputId": "371ec1c2-51b9-48f6-d18c-31d0a0fe2e95"
      },
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluate on test data\n",
            "6/6 [==============================] - 0s 24ms/step - loss: 0.0348 - mean_squared_error: 0.0348 - root_mean_squared_error: 0.1865\n",
            "[0.034794196486473083, 0.034794196486473083, 0.18653202056884766]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# for predict anomalies/outliers in the training dataset\n",
        "prediction = model.predict(test_feats_standardized.values)\n",
        "\n",
        "# for get the mean absolute error between actual and reconstruction/prediction\n",
        "prediction_loss = keras.losses.mae(prediction, test_feats_standardized.values)\n",
        "\n",
        "# for check the prediction loss threshold for 2% of outliers\n",
        "loss_threshold = np.percentile(np.mean(prediction_loss, axis=(1,2)), 98)\n",
        "\n",
        "print(f'The prediction loss threshold for 2% of outliers is {loss_threshold:.2f}')\n",
        "# for visualize the threshold\n",
        "plt.hist(np.mean(prediction_loss, axis=(1,2)), bins=30, alpha=0.8)\n",
        "plt.axvline(x=loss_threshold, color='orange')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 465
        },
        "id": "JF74bbloI-Tz",
        "outputId": "ecc73336-71fe-475b-a3cc-2fe02a4d4b3d"
      },
      "execution_count": 73,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "21/21 [==============================] - 0s 9ms/step\n",
            "The prediction loss threshold for 2% of outliers is 0.27\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# References\n",
        "\n",
        "Campbell, T., G. M. Lackmann, M. J. Molina, and M. D. Parker (2024). Severe Convective Storms in Limited Instability Organized by Pattern and Distribution. Weather and Forecasting. https://journals.ametsoc.org/view/journals/wefo/39/1/WAF-D-23-0130.1.xml\n",
        "\n",
        "DuVivier, A. K., M. J. Molina, A. L. Deppenmeier, M. M. Holland, L. Landrum, K. Krumhardt, and S. Jenouvrier (2023). Projections of Winter Polynyas and Their Biophysical Impacts in the Ross Sea Antarctica. Climate Dynamics. https://link.springer.com/article/10.1007/s00382-023-06951-z\n",
        "\n",
        "Molina, M. J., D. J. Gagne, and A. F. Prein (2021). A benchmark to test generalization capabilities of deep learning methods to classify severe convective storms in a changing climate. Earth and Space Science. https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2020EA001490\n",
        "\n",
        "Molina, M. J., T. A. O'Brien, G. Anderson, M. Ashfaq, K. E. Bennett, W. D. Collins, K. Dagon, J. M. Restrepo, and P. A. Ullrich (2023). A Review of Recent and Emerging Machine Learning Applications for Climate Variability and Weather Phenomena. Artificial Intelligence for the Earth Systems. https://journals.ametsoc.org/view/journals/aies/2/4/AIES-D-22-0086.1.xml\n",
        "\n",
        "Molina, M. J., J. H. Richter, A. A. Glanville, K. Dagon, J. Berner, A. Hu, and G. A. Meehl (2023). Subseasonal Representation and Predictability of North American Weather Regimes using Cluster Analysis. Artificial Intelligence for the Earth Systems. https://journals.ametsoc.org/view/journals/aies/2/2/AIES-D-22-0051.1.xml\n",
        "\n",
        "Passarella, L.S. and Mahajan, S., 2023. Assessing Tropical Pacific–Induced Predictability of Southern California Precipitation Using a Novel Multi-Input Multioutput Autoencoder. Artificial Intelligence for the Earth Systems, 2(4), p.e230003. https://journals.ametsoc.org/view/journals/aies/2/4/AIES-D-23-0003.1.xml"
      ],
      "metadata": {
        "id": "Y8DYzxtePqCH"
      }
    }
  ]
}