eda_rnn.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from tqdm import tqdm\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.io import wavfile\n",
    "from python_speech_features import mfcc, logfbank\n",
    "import librosa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#calculate fft @params signal and rate\n",
    "def calculate_fft(signal, rate):\n",
    "    signal_length = len(signal)\n",
    "    frequency = np.fft.rfftfreq(signal_length, d = 1/rate)\n",
    "    #mean normalization of length of signal\n",
    "    magnitude = abs(np.fft.rfft(signal)/signal_length) \n",
    "    return (magnitude, frequency)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#calculate low and hig frequency because we have dead spots in signal\n",
    "#we do this to get atleats a threshold value of amps\n",
    "def envelope(signal, rate, threshold):\n",
    "    mask = []\n",
    "    signal = pd.Series(signal).apply(np.abs)\n",
    "    signal_mean = signal.rolling(window=int(rate/10), min_periods = 1, center = True).mean()\n",
    "    for mean in signal_mean:\n",
    "        if mean > threshold:\n",
    "            mask.append(True)\n",
    "        else:\n",
    "            mask.append(False)\n",
    "    return mask"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_signals(signals):\n",
    "    fig, axes = plt.subplots(nrows=2, ncols=2, sharex=False,\n",
    "                             sharey=True, figsize=(20,5))\n",
    "    fig.suptitle('Time Series', size=16)\n",
    "    i = 0\n",
    "    for x in range(2):\n",
    "        for y in range(1):\n",
    "            axes[x,y].set_title(list(signals.keys())[i])\n",
    "            axes[x,y].plot(list(signals.values())[i])\n",
    "            axes[x,y].get_xaxis().set_visible(False)\n",
    "            axes[x,y].get_yaxis().set_visible(False)\n",
    "            i += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_fft(fft):\n",
    "    fig, axes = plt.subplots(nrows=2, ncols=2, sharex=False,\n",
    "                             sharey=True, figsize=(20,5))\n",
    "    fig.suptitle('Fourier Transforms', size=16)\n",
    "    i = 0\n",
    "    for x in range(2):\n",
    "        for y in range(1):\n",
    "            data = list(fft.values())[i]\n",
    "            Y, freq = data[0], data[1]\n",
    "            axes[x,y].set_title(list(fft.keys())[i])\n",
    "            axes[x,y].plot(freq, Y)\n",
    "            axes[x,y].get_xaxis().set_visible(False)\n",
    "            axes[x,y].get_yaxis().set_visible(False)\n",
    "            i += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_fbank(fbank):\n",
    "    fig, axes = plt.subplots(nrows=2, ncols=2, sharex=False,\n",
    "                             sharey=True, figsize=(20,5))\n",
    "    fig.suptitle('Filter Bank Coefficients', size=16)\n",
    "    i = 0\n",
    "    for x in range(2):\n",
    "        for y in range(1):\n",
    "            axes[x,y].set_title(list(fbank.keys())[i])\n",
    "            axes[x,y].imshow(list(fbank.values())[i],\n",
    "                    cmap='hot', interpolation='nearest')\n",
    "            axes[x,y].get_xaxis().set_visible(False)\n",
    "            axes[x,y].get_yaxis().set_visible(False)\n",
    "            i += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_mfccs(mfccs):\n",
    "    fig, axes = plt.subplots(nrows=2, ncols=2, sharex=False,\n",
    "                             sharey=True, figsize=(20,5))\n",
    "    fig.suptitle('Mel Frequency Cepstrum Coefficients', size=16)\n",
    "    i = 0\n",
    "    for x in range(2):\n",
    "        for y in range(1):\n",
    "            axes[x,y].set_title(list(mfccs.keys())[i])\n",
    "            axes[x,y].imshow(list(mfccs.values())[i],\n",
    "                    cmap='hot', interpolation='nearest')\n",
    "            axes[x,y].get_xaxis().set_visible(False)\n",
    "            axes[x,y].get_yaxis().set_visible(False)\n",
    "            i += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extract files with labels\n",
      "neg-0421-083-cough-m-53-0.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-0.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-1.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-1.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-10.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-10.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-11.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-11.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-12.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-12.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-13.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-13.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-14.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-14.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-15.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-15.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-16.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-16.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-17.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-17.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-2.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-2.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-3.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-3.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-4.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-4.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-5.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-5.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-6.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-6.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-7.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-7.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-8.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-8.wav, dtype: object\n",
      "neg-0421-083-cough-m-53-9.wav corona_test    negative\n",
      "Name: neg-0421-083-cough-m-53-9.wav, dtype: object\n",
      "neg-0421-085-cough-m-43-1.wav corona_test    negative\n",
      "Name: neg-0421-085-cough-m-43-1.wav, dtype: object\n",
      "neg-0421-085-cough-m-43-2.wav corona_test    negative\n",
      "Name: neg-0421-085-cough-m-43-2.wav, dtype: object\n",
      "neg-0421-088-cough-f-66-0.wav corona_test    negative\n",
      "Name: neg-0421-088-cough-f-66-0.wav, dtype: object\n",
      "neg-0421-088-cough-f-66-1.wav corona_test    negative\n",
      "Name: neg-0421-088-cough-f-66-1.wav, dtype: object\n",
      "neg-0421-088-cough-f-66-2.wav corona_test    negative\n",
      "Name: neg-0421-088-cough-f-66-2.wav, dtype: object\n",
      "neg-0421-088-cough-f-66-3.wav corona_test    negative\n",
      "Name: neg-0421-088-cough-f-66-3.wav, dtype: object\n",
      "neg-0421-088-cough-f-66-4.wav corona_test    negative\n",
      "Name: neg-0421-088-cough-f-66-4.wav, dtype: object\n",
      "neg-0421-088-cough-f-66-5.wav corona_test    negative\n",
      "Name: neg-0421-088-cough-f-66-5.wav, dtype: object\n",
      "neg-0421-088-cough-f-66-6.wav corona_test    negative\n",
      "Name: neg-0421-088-cough-f-66-6.wav, dtype: object\n",
      "neg-0421-088-cough-f-66-8.wav corona_test    negative\n",
      "Name: neg-0421-088-cough-f-66-8.wav, dtype: object\n",
      "neg-0421-089-cough-f-20-0.wav corona_test    negative\n",
      "Name: neg-0421-089-cough-f-20-0.wav, dtype: object\n",
      "neg-0421-089-cough-f-20-2.wav corona_test    negative\n",
      "Name: neg-0421-089-cough-f-20-2.wav, dtype: object\n",
      "neg-0421-089-cough-f-20-3.wav corona_test    negative\n",
      "Name: neg-0421-089-cough-f-20-3.wav, dtype: object\n",
      "neg-0421-089-cough-f-20-4.wav corona_test    negative\n",
      "Name: neg-0421-089-cough-f-20-4.wav, dtype: object\n",
      "neg-0421-089-cough-f-20-5.wav corona_test    negative\n",
      "Name: neg-0421-089-cough-f-20-5.wav, dtype: object\n",
      "neg-0421-089-cough-f-20-6.wav corona_test    negative\n",
      "Name: neg-0421-089-cough-f-20-6.wav, dtype: object\n",
      "neg-0421-090-cough-f-17-1.wav corona_test    negative\n",
      "Name: neg-0421-090-cough-f-17-1.wav, dtype: object\n",
      "neg-0421-090-cough-f-17-11.wav corona_test    negative\n",
      "Name: neg-0421-090-cough-f-17-11.wav, dtype: object\n",
      "neg-0421-090-cough-f-17-2.wav corona_test    negative\n",
      "Name: neg-0421-090-cough-f-17-2.wav, dtype: object\n",
      "neg-0421-090-cough-f-17-3.wav corona_test    negative\n",
      "Name: neg-0421-090-cough-f-17-3.wav, dtype: object\n",
      "neg-0421-090-cough-f-17-4.wav corona_test    negative\n",
      "Name: neg-0421-090-cough-f-17-4.wav, dtype: object\n",
      "neg-0421-090-cough-f-17-5.wav corona_test    negative\n",
      "Name: neg-0421-090-cough-f-17-5.wav, dtype: object\n",
      "neg-0421-090-cough-f-17-6.wav corona_test    negative\n",
      "Name: neg-0421-090-cough-f-17-6.wav, dtype: object\n",
      "neg-0421-090-cough-f-17-7.wav corona_test    negative\n",
      "Name: neg-0421-090-cough-f-17-7.wav, dtype: object\n",
      "neg-0421-091-cough-m-47-1.wav corona_test    negative\n",
      "Name: neg-0421-091-cough-m-47-1.wav, dtype: object\n",
      "neg-0421-091-cough-m-47-2.wav corona_test    negative\n",
      "Name: neg-0421-091-cough-m-47-2.wav, dtype: object\n",
      "neg-0421-091-cough-m-47-3.wav corona_test    negative\n",
      "Name: neg-0421-091-cough-m-47-3.wav, dtype: object\n",
      "neg-0421-091-cough-m-47-4.wav corona_test    negative\n",
      "Name: neg-0421-091-cough-m-47-4.wav, dtype: object\n",
      "neg-0421-091-cough-m-47-5.wav corona_test    negative\n",
      "Name: neg-0421-091-cough-m-47-5.wav, dtype: object\n",
      "neg-0421-091-cough-m-47-6.wav corona_test    negative\n",
      "Name: neg-0421-091-cough-m-47-6.wav, dtype: object\n",
      "neg-0421-091-cough-m-47-7.wav corona_test    negative\n",
      "Name: neg-0421-091-cough-m-47-7.wav, dtype: object\n",
      "neg-0421-091-cough-m-47-8.wav corona_test    negative\n",
      "Name: neg-0421-091-cough-m-47-8.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-0.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-0.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-1.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-1.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-10.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-10.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-12.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-12.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-13.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-13.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-14.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-14.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-15.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-15.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-2.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-2.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-3.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-3.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-4.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-4.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-5.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-5.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-6.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-6.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-7.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-7.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-8.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-8.wav, dtype: object\n",
      "neg-0422-095-cough-m-53-9.wav corona_test    negative\n",
      "Name: neg-0422-095-cough-m-53-9.wav, dtype: object\n",
      "neg-0422-097-cough-m-37-1.wav corona_test    negative\n",
      "Name: neg-0422-097-cough-m-37-1.wav, dtype: object\n",
      "neg-0422-097-cough-m-37-3.wav corona_test    negative\n",
      "Name: neg-0422-097-cough-m-37-3.wav, dtype: object\n",
      "neg-0422-097-cough-m-37-4.wav corona_test    negative\n",
      "Name: neg-0422-097-cough-m-37-4.wav, dtype: object\n",
      "neg-0422-097-cough-m-37-8.wav corona_test    negative\n",
      "Name: neg-0422-097-cough-m-37-8.wav, dtype: object\n",
      "neg-0422-097-cough-m-37-9.wav corona_test    negative\n",
      "Name: neg-0422-097-cough-m-37-9.wav, dtype: object\n",
      "neg-0422-098-cough-f-24-0.wav corona_test    negative\n",
      "Name: neg-0422-098-cough-f-24-0.wav, dtype: object\n",
      "neg-0422-098-cough-f-24-1.wav corona_test    negative\n",
      "Name: neg-0422-098-cough-f-24-1.wav, dtype: object\n",
      "neg-0422-098-cough-f-24-5.wav corona_test    negative\n",
      "Name: neg-0422-098-cough-f-24-5.wav, dtype: object\n",
      "pos-0421-084-cough-m-50-0.wav corona_test    positive\n",
      "Name: pos-0421-084-cough-m-50-0.wav, dtype: object\n",
      "pos-0421-084-cough-m-50-1.wav corona_test    positive\n",
      "Name: pos-0421-084-cough-m-50-1.wav, dtype: object\n",
      "pos-0421-084-cough-m-50-2.wav corona_test    positive\n",
      "Name: pos-0421-084-cough-m-50-2.wav, dtype: object\n",
      "pos-0421-084-cough-m-50-3.wav corona_test    positive\n",
      "Name: pos-0421-084-cough-m-50-3.wav, dtype: object\n",
      "pos-0421-084-cough-m-50-4.wav corona_test    positive\n",
      "Name: pos-0421-084-cough-m-50-4.wav, dtype: object\n",
      "pos-0421-084-cough-m-50-5.wav corona_test    positive\n",
      "Name: pos-0421-084-cough-m-50-5.wav, dtype: object\n",
      "pos-0421-084-cough-m-50-6.wav corona_test    positive\n",
      "Name: pos-0421-084-cough-m-50-6.wav, dtype: object\n",
      "pos-0421-084-cough-m-50-7.wav corona_test    positive\n",
      "Name: pos-0421-084-cough-m-50-7.wav, dtype: object\n",
      "pos-0421-086-cough-m-65-0.wav corona_test    positive\n",
      "Name: pos-0421-086-cough-m-65-0.wav, dtype: object\n",
      "pos-0421-086-cough-m-65-1.wav corona_test    positive\n",
      "Name: pos-0421-086-cough-m-65-1.wav, dtype: object\n",
      "pos-0421-086-cough-m-65-2.wav corona_test    positive\n",
      "Name: pos-0421-086-cough-m-65-2.wav, dtype: object\n",
      "pos-0421-086-cough-m-65-3.wav corona_test    positive\n",
      "Name: pos-0421-086-cough-m-65-3.wav, dtype: object\n",
      "pos-0421-086-cough-m-65-4.wav corona_test    positive\n",
      "Name: pos-0421-086-cough-m-65-4.wav, dtype: object\n",
      "pos-0421-086-cough-m-65-5.wav corona_test    positive\n",
      "Name: pos-0421-086-cough-m-65-5.wav, dtype: object\n",
      "pos-0421-086-cough-m-65-6.wav corona_test    positive\n",
      "Name: pos-0421-086-cough-m-65-6.wav, dtype: object\n",
      "pos-0421-086-cough-m-65-7.wav corona_test    positive\n",
      "Name: pos-0421-086-cough-m-65-7.wav, dtype: object\n",
      "pos-0421-086-cough-m-65-8.wav corona_test    positive\n",
      "Name: pos-0421-086-cough-m-65-8.wav, dtype: object\n",
      "pos-0421-087-cough-f-40-0.wav corona_test    positive\n",
      "Name: pos-0421-087-cough-f-40-0.wav, dtype: object\n",
      "pos-0421-087-cough-f-40-1.wav corona_test    positive\n",
      "Name: pos-0421-087-cough-f-40-1.wav, dtype: object\n",
      "pos-0421-087-cough-f-40-2.wav corona_test    positive\n",
      "Name: pos-0421-087-cough-f-40-2.wav, dtype: object\n",
      "pos-0421-087-cough-f-40-3.wav corona_test    positive\n",
      "Name: pos-0421-087-cough-f-40-3.wav, dtype: object\n",
      "pos-0421-087-cough-f-40-5.wav corona_test    positive\n",
      "Name: pos-0421-087-cough-f-40-5.wav, dtype: object\n",
      "pos-0421-087-cough-f-40-6.wav corona_test    positive\n",
      "Name: pos-0421-087-cough-f-40-6.wav, dtype: object\n",
      "pos-0421-092-cough-m-53-0.wav corona_test    positive\n",
      "Name: pos-0421-092-cough-m-53-0.wav, dtype: object\n",
      "pos-0421-092-cough-m-53-1.wav corona_test    positive\n",
      "Name: pos-0421-092-cough-m-53-1.wav, dtype: object\n",
      "pos-0421-092-cough-m-53-12.wav"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 0/2 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " corona_test    positive\n",
      "Name: pos-0421-092-cough-m-53-12.wav, dtype: object\n",
      "pos-0421-092-cough-m-53-13.wav corona_test    positive\n",
      "Name: pos-0421-092-cough-m-53-13.wav, dtype: object\n",
      "pos-0421-092-cough-m-53-3.wav corona_test    positive\n",
      "Name: pos-0421-092-cough-m-53-3.wav, dtype: object\n",
      "pos-0421-092-cough-m-53-5.wav corona_test    positive\n",
      "Name: pos-0421-092-cough-m-53-5.wav, dtype: object\n",
      "pos-0421-092-cough-m-53-6.wav corona_test    positive\n",
      "Name: pos-0421-092-cough-m-53-6.wav, dtype: object\n",
      "pos-0421-092-cough-m-53-7.wav corona_test    positive\n",
      "Name: pos-0421-092-cough-m-53-7.wav, dtype: object\n",
      "pos-0421-092-cough-m-53-9.wav corona_test    positive\n",
      "Name: pos-0421-092-cough-m-53-9.wav, dtype: object\n",
      "pos-0421-093-cough-f-24-0.wav corona_test    positive\n",
      "Name: pos-0421-093-cough-f-24-0.wav, dtype: object\n",
      "pos-0421-093-cough-f-24-1.wav corona_test    positive\n",
      "Name: pos-0421-093-cough-f-24-1.wav, dtype: object\n",
      "pos-0421-093-cough-f-24-2.wav corona_test    positive\n",
      "Name: pos-0421-093-cough-f-24-2.wav, dtype: object\n",
      "pos-0421-093-cough-f-24-5.wav corona_test    positive\n",
      "Name: pos-0421-093-cough-f-24-5.wav, dtype: object\n",
      "pos-0421-093-cough-f-24-6.wav corona_test    positive\n",
      "Name: pos-0421-093-cough-f-24-6.wav, dtype: object\n",
      "pos-0421-094-cough-m-51-0.wav corona_test    positive\n",
      "Name: pos-0421-094-cough-m-51-0.wav, dtype: object\n",
      "pos-0421-094-cough-m-51-2.wav corona_test    positive\n",
      "Name: pos-0421-094-cough-m-51-2.wav, dtype: object\n",
      "pos-0421-094-cough-m-51-3.wav corona_test    positive\n",
      "Name: pos-0421-094-cough-m-51-3.wav, dtype: object\n",
      "pos-0421-094-cough-m-51-4.wav corona_test    positive\n",
      "Name: pos-0421-094-cough-m-51-4.wav, dtype: object\n",
      "pos-0421-094-cough-m-51-5.wav corona_test    positive\n",
      "Name: pos-0421-094-cough-m-51-5.wav, dtype: object\n",
      "pos-0421-094-cough-m-51-6.wav corona_test    positive\n",
      "Name: pos-0421-094-cough-m-51-6.wav, dtype: object\n",
      "pos-0421-094-cough-m-51-7.wav corona_test    positive\n",
      "Name: pos-0421-094-cough-m-51-7.wav, dtype: object\n",
      "pos-0422-096-cough-m-31-1.wav corona_test    positive\n",
      "Name: pos-0422-096-cough-m-31-1.wav, dtype: object\n",
      "pos-0422-096-cough-m-31-2.wav corona_test    positive\n",
      "Name: pos-0422-096-cough-m-31-2.wav, dtype: object\n",
      "pos-0422-096-cough-m-31-6.wav corona_test    positive\n",
      "Name: pos-0422-096-cough-m-31-6.wav, dtype: object\n",
      "pos-0422-096-cough-m-31-8.wav corona_test    positive\n",
      "Name: pos-0422-096-cough-m-31-8.wav, dtype: object\n",
      "['negative', 'positive']\n",
      "Reading audio files with librosa\n",
      "neg-0421-083-cough-m-53-0.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 50%|█████     | 1/2 [00:02<00:02,  2.26s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "pos-0421-084-cough-m-50-0.wav\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2/2 [00:02<00:00,  1.25s/it]\n"
     ]
    }
   ],
   "source": [
    "## file operations\n",
    "\n",
    "##read file labels\n",
    "data_frame = pd.read_csv('dataset_minor2.csv')\n",
    "data_frame.set_index('cough_filename', inplace=True)\n",
    "\n",
    "print('Extract files with labels')\n",
    "for index, row in data_frame.iterrows():\n",
    "    print(index,row)\n",
    "    if row['corona_test']=='negative':\n",
    "        rate, signal = wavfile.read('./clinical/converted_seg/cneg/'+index)\n",
    "        data_frame.at[index, 'length'] = signal.shape[0]/rate\n",
    "    else:\n",
    "        rate, signal = wavfile.read('./clinical/converted_seg/cpos/'+index)\n",
    "        data_frame.at[index, 'length'] = signal.shape[0]/rate\n",
    "\n",
    "\n",
    "classes = list(np.unique(data_frame.corona_test))\n",
    "class_dist = data_frame.groupby(['corona_test'])['length'].mean()\n",
    "print(classes)\n",
    "\n",
    "\n",
    "data_frame.reset_index(inplace=True)\n",
    "\n",
    "##datapoints\n",
    "signals = {}\n",
    "fft = {}\n",
    "fbank = {}\n",
    "mfccs = {}\n",
    "\n",
    "print(\"Reading audio files with librosa\")\n",
    "for c in tqdm(classes):\n",
    "    wav_file = data_frame[data_frame.corona_test == c].iloc[0,0]\n",
    "    print(wav_file)\n",
    "    if c == 'negative':\n",
    "        signal, rate = librosa.load('clinical/converted_seg/cneg/'+ wav_file, sr=44100)\n",
    "    else:\n",
    "        signal, rate = librosa.load('clinical/converted_seg/cpos/'+ wav_file, sr=44100)\n",
    "    \n",
    "    mask = envelope(signal, rate, 0.0005)\n",
    "    signal = signal[mask]\n",
    "    signals[c] = signal\n",
    "    fft[c] = calculate_fft(signal, rate)\n",
    "\n",
    "    bank = logfbank(signal[:rate], rate, nfilt=26, nfft=1103).T\n",
    "    fbank[c] = bank\n",
    "    \n",
    "    mel = mfcc(signal[:rate], rate, numcep=40, nfilt = 26, nfft = 1103).T\n",
    "    mfccs[c] = mel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Writing clean files for calssification training\n"
     ]
    }
   ],
   "source": [
    "#writing cleaned files for training\n",
    "\n",
    "print(\"Writing clean files for calssification training\")\n",
    "for index, row in data_frame.iterrows():\n",
    "#     for f in tqdm(data_frame.cough_filename):\n",
    "    try:\n",
    "        if(row['corona_test']=='negative'):\n",
    "            signal, rate = librosa.load('./clinical/converted_seg/cneg/'+row[\"cough_filename\"], sr = 16000)\n",
    "        else:\n",
    "            signal, rate = librosa.load('./clinical/converted_seg/cpos/'+row[\"cough_filename\"], sr = 16000)\n",
    "        mask = envelope(signal, rate, 0.0005)\n",
    "        wavfile.write(filename='cleanfiles/'+row[\"cough_filename\"], rate = rate, data= signal[mask])\n",
    "    except FileNotFoundError:\n",
    "        print(\"Unexpected error while reading files:\", arg)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# #writing files for testing\n",
    "# print(os.listdir('testcleanfiles'))\n",
    "\n",
    "# print(\"Writing clean files for calssification\")\n",
    "# for f in tqdm(data_frame.cough_filename):\n",
    "#     try:\n",
    "#         signal, rate = librosa.load('testsounds/'+f, sr = 16000)\n",
    "#         mask = envelope(signal, rate, 0.0005)\n",
    "#         wavfile.write(filename='testcleanfiles/'+f, rate = rate, data= signal[mask])\n",
    "#     except FileNotFoundError:\n",
    "#         print(\"Unexpected error while reading files:\", arg)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot class distribution\n",
    "fig, ax = plt.subplots()\n",
    "ax.set_title(\"class distribution\", y=1.08)\n",
    "ax.pie(class_dist, labels=class_dist.index, autopct='%1.1f%%', shadow=False, startangle=90)\n",
    "ax.axis('equal')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot raw signal\n",
    "plot_signals(signals)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot fft signals\n",
    "plot_fft(fft)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot Filter bank energy\n",
    "plot_fbank(fbank)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA/YAAAFTCAYAAACTVpcgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3debxdZXn//e9lQogJCSHJAQQCEQUsilA4CFIRHBGlUpyqotShxbn91bY/tE48Ra0+fdpapA5UgTrhUEQRRQQUoSLKgTIrU0wIQyAJBEJiBpLr+WPtlH3u9T1knU3OsHI+79frvJL7Otda+17rXicn917rvnZkpgAAAAAAQDs9aaw7AAAAAAAAesfEHgAAAACAFmNiDwAAAABAizGxBwAAAACgxZjYAwAAAADQYkzsAQAAAABoMSb2AJ6wiHhLRGTna2/z/SO7vv/iHvZ/aURc2iAvh/j62nBfE15Ujo+ISyJieUSsj4i7IuKbEfGCMezXARFxckTMHqs+9CIido+I0yLitohYExGPRMRVEfGhiNh+BF/3bZ3XXBcRKzqx6RHx1Yi4v/Nz85mun90jh7n/jIiTR6Lvnf0f2Rlv/h8DAICkyWPdAQBblZWS3izpI0X8hM73ZoxCH86S9MUitnQUXnerFxGTJH1T0nGS/lPSZyU9IGmepNdKuiQidsjMh8agewdI+pikr3X6NO5FxPMlnSfpfkmnSrpR0jaSDpX0HklzJf31CLzuLpJOl/R1SW+VtKbzrfdIeoOkt0m6VdK9kh6U9FxJNw/zZZ4r6a4t0d8hHKlqvD8uaeMIvg4AAK3AxB7AlvRdSW+KiI9mZkpSRDxZ0qslnSPpLaPQh7sz88omiZ2JamTmoyPcp63FByW9RtJrMvOc4ntfj4iXSlo/+t0anvEw7hGxg6T/kvQbSS/OzFVd3/5JRPyzpMNG6OX3kjRJ0n9m5n93xf9A0j2Z+ZUiv9HPU7emP4MAAGDL4BE2AFvSVyXtIel5XbHjVE0iyomgJCkijug81r0yIlZFxIUR8ayR6Fzn8eBPRMQHIuJ3ktZJ2q9pPyJiUkR8PCLujYjVnSUCzywfO46IsyJioXn92pKCiJgbEZ+PiLsjYm1E/DYiTixyNi11ODQivh4RD0fEPRFxakRMLXKnR8SnIuKOzv6WRMQ5EbFTRBzU2c+xpm9ndR6pnzTEuZsi6W8k/dBM6iVJmfmTzFzdtU2Tc3ppRPx3RBwbETd2nYPXFXl7R8S5ncfE10TEnRHxnYiYHBFvkXRmJ/W2eGwJxvzOtnbcu87r/OK1To6ILGLZGfu/iYhFneP5YUTs2Pn6dkQ8FBGLI+Ikd34KfyGpT9L7ikn9pnO5KjMv6nr9p0TEVyJiWeccXR8Rbyq3i4indq6RpZ28ayPiuK7vnyXp0k7zks5xndU53rdImtd1/o6MIR7Fj4jjIuIXUS0deDgifh0RryzO18nFNvtHxHkR8WBE/L6z/eFFzqbr8A8j4vLOz9ltEfHOrpyTVd2tl6T1m/rb+d7kiDilc/2v6Zyv/46I7n+TAADY6jCxB7AlLZJ0marH8Tc5QdK5kh4pkyPiFZIu6XzvTZLeqOpx/csjYl6PfYjOf+7/96v4/lskvULS33b+vGcY/ThZ0t+reoT5TyT9RNWj1L12dKakX3T6cXLnzx9I+nxEvM9s8lVJd0h6laTPq3p0+oNd+5si6SJJf6lqScIxkt6r6tH0HTLzaklXSXpH0Y9Zkl4n6UuZuWGI7vZLmqWGxzvMsX26qkfR/7lzbLdLKtfsny9pV0nvknSUpA9IWqvq99gPVT2SLVVLAp7b+bq3a/u3qBj3JsdReLOkF0p6t6T3STpc0ldUXd/Xq3oy5UeSPhURL9/Mvl4saUlmDmzuRSNiuqSfSzpa1fX3J5JukPTV6HoTqHNefyVpf1WP8L9S0jWSzumadJ+i6vqQquvnuZ3YcyVdKGmJHjt/1wzRn/epejrnfkl/puqcnytp/uMcw4GSrpA0W9WbGq+WtFzSxRFxUJE+U9I3VC2rOFbVNfv5ruvhS5K+3Pn787r6K0kndY79VFXXyVtVXYetqr0AAMCwZSZffPHF1xP6UjVpSlUTtLepWpc7VdJTJD0q6SWq1sSmqseON213u6RLin3NlLRM0me6YpdKurRBP3KIr6d3ff8eSU8utttsPyTtoGqS+oUi76TOfk/uip0laaHp36DjUFWLYI2kvYq8/+i89uTi/P4/Rd75km7tar+tk/fKzYzVBkl7dMX+sjNOuz3Odn/a2fdRDa+J4YxtSjq0KzZJ0m8lXd5pz214XP871ua6cOO+aZv5RfxkSWn2ceumMenE/qUT/3BXbLKqCe+Zmzk/v5H0y4bn8r2d1zmyiF/cea1JnfaXVdWTmFPkXSTp2q72i4fY39fK61aP/dwe2TWGKyV9t8HPYvfPxCWdY55SjPNvJH2v+NlJSS/oim3buW5OL8eoezy6fiYet2988cUXX3zxtTV+cccewJb2HVX/Ef9jScerugN4SZkUEXtJepqqtdndd9dXS/qlpOf3+PpnSDq4+Frc9f0fZ+bve+jHfpKmS/p28Xrf7LGfkvQyVXdYf1e89oWS5kjat8j/YdG+QdLuXe2XqroL/Hh31b8paYWqu6abvEPVI/ZbpNhZD2O7OLvWZGf11MB3JD0nqqrnyyUtUHUn/C86+x+uQePeo4ty8Lr833b+vHBToPP921UVFNxSnq+qdsSlRfxrqh7n33SdvEzVEwMPmetp/84TIk/UYZK2U1V8r5Go6mwcoWpMN3b1K1S9OVFeD6sz82ebGpm5VtJtGnytD+UqSS/vLL14XucpFgAAtnpM7AFsUZm5UtL3VD22fIKkr2emq1q9Y+fPL6squNb9dYyqiW0v7s3MgeJrbff3e+zHUzp/3ldsX7aHY0dVk5rydb/T+X55Dspq72tVvYmyyRxJdz/eC2bmGlXr0d/emWAdrmpi+IXN9HXTmyN7bCZPGv7YunN4n6QpkvoyM1U99TEg6R8l3RoRCyLiXQ36skk57r14sGive5z4VD2+xXqcR9cLs+X7v6Tr+1J13k9Q/Zz/U+f7vf5Mddu0j+G8CTRb1d35j5i+vVfSDjH4Y+vK8ylV1/rmzqkkfVLV+vtXSrpc0vKIODMi5g6jvwAAtA5V8QGMhK+ourv8JFUfn+Us7/z5QVV37UrrTGxLyKLdtB+bJlY7Sbqp6/s7mW3WqJqUluZ0vd6m175f0l8N0ddbhogPZZmkJoUHPy/p/arWLx8naaG67joPYUDVnf4/1ubv1g53bN053KmTt1SSMnOBpBMiIlStIX+vpM9FxMLMvGAz/ZHq4y499jFv5VhtiQnw5lws6SURcVBWtQ8ezwOS9jHxnTt/Lu/683JJnx5iP73UFSgt6/y5q6qP52tihaqPpPt3Vf821Azx5t+wZeZ6Vcf/6YjYWdUbSf8iaZqq5SQAAGyVmNgDGAkXqXpkfUVm3jREzi2qJpTPzMxPjVbHnkA/rpe0SlWRuZ92xV9vchdJ2iki5mbmMkmKiKepmpxd0ZX3Y1VF2O7MzPt7PoLH/ETS6yPijzPzB0MlZeYdEfETSX+n6vPf/2FzE6vMXBfVR7CdEhGvTlMZPyJeoqoY4HDHdl5EHLrpcfyoKvO/VtKvy3517t5fGxHvl/R2VW9kXKDqjq4kPbnB622yqPPns1Stn1fnEfGXDmMfvfqSqvN/WkSUH3eniJgm6bDMvFhV4bzXRsQfZeYvutLeqOqNod902j9WVUTupi2w7GAoV6iqNXGiNv9mkKSqwn9EXK7qDZlrttAkvnu8Vw7xukskfalTyHBEPmkDAIDxgok9gC2us0Z6qDv1m3IyIt4j6fuddbDfVnU3cCdV63jvzMx/GYW+NupHZq6IiH+V9KGIWKlqEn2wqsll6TuqKo1/PSL+RVXxtw/qsbudm/yrqruIl3f2fYuqdfzPkHR4ZtY+lm4zvqZq7fzZEfGPqtbvz1BVHfwzmfnbrtzPSfq+qsehz2i4/39UNTn7VlQfm/YDVXeTd1NV5fxVqqrvrx7m2N7X2efHVN2hf5ekvTt/KiKeLenfJH1L1fr1SaoK3z2qx95kubnz53si4j87x3V9Zj7ekx9XqfqUgX/qPAq+VlXF+20fZ5stIjMfiIhXq/qUgWsi4rOq7oBvI+k5kt6p6nPuL1ZVUO6vJH03Ij6k6jH441UtT3hHPvZJBh+V9GtJl0XEaareXNlB1aR2z8x82xbo98qI+KCkz0bEOao+IWKlqjeI1mTmZ4fY9P2qPjHjwoj4sqonYOZKOlBV8b8PDLMrm8b7byLiAkkbMnMgIr4v6TpVFf0flPSHqmoPfHGY+wcAoFWY2AMYM5n5o4h4vqQPqbqD+WRV64avVDWJG2/9OFlVwa8/V/Uo+K9UPZp+U7G/2yPiNao+gu17qu4Gv1/VR5V15z0UEYepmpCdpOrx5hWqJvj2s+I3cxzrI+KlqtYYn9j5c7mqu+jl+vwfqipm96POnc0m+98Q1efLH6/qDY2zVBVSu0/VI+BHZOZDndzhjO3tkv5fVeuj91I1IX1DVwG1JZLuVHUOd1P1CP0Nko7Z9Bh7Zl7X+XzzE1W9ufEkSU/t7Guo43k0Io5V9Yj4WarO0WdUjevHhtpuS8nMyyJif1V37v9a1bGtV3UH/t9Vvfmy6Y73EarO0adUvVlzi6Q3Z+bXuvZ3Z0T0q7pOP6mqsN5yVW8Y/OcW7PdpEbGk0++vd/X5lMfZ5pqIOFjVeT1V0vaq3sS5Rpuv7+Ccr+r8vFvVz090vi5T9bTHe1Q9fn+nqvP2iR5eAwCA1ojqqUYAQK8iYtNH0Z081n1pqvPY/E9Uffxg7VMLRrEfl6r6yLLnjVUfAAAA2o479gAwgXTW+u+pahnANWM5qQcAAMCWwcfdAcDE8hE9VmzuhDHuCwAAALYAHsUHAAAAAKDFuGMPAAAAAECLMbEHAAAAAKDFmNgDAAAAANBiTOwBAAAAAGgxJvZ4wiLikYjYc6z7AQAAAAATERN7DEtEXBoRf94dy8ztMnPBWPUJAAAAACYyJvYAAAAAALQYE/sWi4iFEfG3EXF9RDwUEd+KiKmd7x0TEddGxIqIuCIint213YER8T8RsTIivtPZ7uOd7+0QEedHxNKIeLDz99063/uEpMMlndZ5/P60Tjwj4ukRcWhELImISV2vdVxEXN/5+5Mi4gMRcUdELI+Ib0fE7NE7YwAAAACw9WFi336vk/QySU+V9GxJb4mIAyWdIekdkuZI+qKk8yJi24iYIulcSWdJmi3pbEnHde3vSZLOlLSHpN0l/V7SaZKUmR+SdLmk93Yev39vd0cy80pJqyS9sCv8Rknf6Pz9LyX9iaQjJO0i6UFJ//6EzwAAAAAATGBM7Nvv1My8JzMfkPQDSQdI+gtJX8zMX2Xmhsz8T0lrJR3a+Zrc2W59Zn5X0q837Swzl2fmOZm5OjNXSvqEqol4U2dLeoMkRcQMSS/vxKTqjYYPZeZdmblW0smSXhMRk3s+egAAAACY4JjYt9+Srr+vlrSdqrvtf9N5DH9FRKyQNE/VXfJdJN2dmdm13eJNf4mIaRHxxYhYFBEPS7pM0qzux+s34xuSXhUR20p6laRrMnNR53t7SDq3q0+/kbRB0k7DPWgAAAAAQIWJ/dZpsaRPZOasrq9pmXm2pHsl7RoR0ZU/r+vvfyNpH0mHZOZMSc/vxDfld78hUJOZN0taJOloDX4Mf1O/ji76NTUz7+71QAEAAABgomNiv3X6D0nvjIhDojI9Il7ReTT+l6rukr83IiZHxLGSntO17QxV6+pXdArbfazY932SNveZ9d9QtZ7++ZK+0xX/gqRPRMQekhQRfZ3XBwAAAAD0iIn9VigzB1Stsz9NVYG62yW9pfO9daoekX+7pBWS3iTpfFVr8CXpM5KeLGmZpCsl/bjY/b+pWhf/YEScOkQXzpZ0pKSfZuayYtvzJP0kIlZ29n9Ir8cJAAAAAJBi8FJrTEQR8StJX8jMM8e6LwAAAACA4eGO/QQUEUdExM6dR/H/TNXH5JV35gEAAAAALcDHjE1M+0j6tqoK+ndIek1m3ju2XQIAAAAA9IJH8QEAAAAAaDEexQcAAAAAoMWY2AMAAAAA0GKtWWM/d0bk/DlbaGcjvfqg6f577Uf0uF1TZb+29PE0yXM5YzFuI3kuXM7Gwc2Fq6Vla3OkRxxAC8ydOzfnz58/1t0AAAAj5Oqrr16WmX29bNuaif38OdLAR4vghh535rbbaGK9arr/Xvs/qcftmj6fsb5oP5Hj6fW4yz5I0roG+296jE37VfbD9ctt1+RcuJw1g5v9l5gcABPS/PnzNTAwMNbdAAAAIyQiFvW6LY/iAwAwTkXEiRExEBEDS5cuHevuAACAcYqJPQAA41Rmnp6Z/ZnZ39fX05N5AABgAmBiDwAAAABAi7Vmjb22lbRHEXvRzHreLx4e3P6jv2v4Aj+sh3538+C2K94387B67N4r6jG3NrtcK7/r803Sq0zs5yb2aNE+w+ScamLXmtgrivbUBq8nSRfWQ+u+U49NeXYRuN3s6ygTW1MPPXjB4PYOf2y2m29ivzCxG+uhR4uF/ZPfZbYz4/HgzfXYiqLtrolfFe2rTQ4AAAAAdOGOPQAAAAAALcbEHgAAAACAFmNiDwAAAABAizGxBwAAAACgxSIzx7oPjcyLyL8uYl82eR8t2u8zOatM7GATe1bRdp8gfL6JvdzE3IcUzSja3zA5D5jYKQ1e82yTc4yJLTGxshScq/HmyuntYmLPM7F7ivY2JucCE1tpYnsW7dtMjim5V9tOqtcylKR9i3ZZ207yY+ti1zTIKV/vw5IWZIZJBTDB9Pf358DAwFh3AwAAjJCIuDoz+3vZljv2AAAAAAC0GBN7AAAAAABajIk9AAAAAAAt1po19v0HRQ78sghOebPJPKdon2FyVtRD695Zj/2gaM8zu3rOlHrspnUm0SgX2e8+zSQdYmJ3mFi56n3XespNP6vHFptdlas6ppsct3D9ZhNbZGJPL9oPmZwXm5hbZH9D0d7f5Gxn3r+6c2M95hbQlwUZnmFyXPGFDSZWKsdfklYPbva/Xxq4jTX2AFhjDwDA1o419gAAAAAATFBM7AEAAAAAaDEm9gAAAAAAtBgTewAAAAAAWmzyWHegsdhWmrLH4NitX63n7VK0F72+nvNbs/8XmlhZX+0Wk7PaFMpzReSWmNh9RXvN6nrOnqbg3dFmX2XtvL5b6zmuWNseJnZ+0XbF4Q43MVdQb3sTKw/T9Wu5ibl+7Fu0XaE/HVUP7f6Lemzlw/VYX9E2dRe1oGGsrI24j8k5u2g/YHIAAAAAoAt37AEAAAAAaDEm9gAAAAAAtBgTewAAAAAAWoyJPQAAAAAALdae4nm3r5WOKQrCmVpzumpwc/kj9ZTZZrNrTGxe0Xa14Rab2M4m5mqglftfaHL2M7Hp3zLBSUX7NpPzhgadkKSfF21TtG7d/63HpvSbfe1kYrOKdlnwUJIWmdj3Taw8pl1NzqoL6rGyMKIk3WFiBxTtM02OK/S3p4kV5zGX1VPieLMdAAAAADwO7tgDAAAAANBiTOwBABinIuLEiBiIiIGlS91nfgIAADCxBwBg3MrM0zOzPzP7+/r6xro7AABgnGJiDwAAAABAi7WneN58SWcUMfe2xJcHN+e4SnlloTlJB73O5N1StM3Nkt13f0E9eOvPaqGnuWJwRd92djdjdn9Jg45J+t2dg9uusODlJjbFxA4v2uY8T5lqtlthYutN7MAG2z17bxM8qB5advbg9lwzHlpbD11/RT12s9l0TtH++GdM0t0mNmBizxrUClvp7xeDm/2XmRwAAAAAeAx37AEAAAAAaDEm9gAAAAAAtBgTewAAAAAAWiwyc6z70Ej/UyIH3l4ErzWJ5ZroaSbHrQ836+41r2i7TxpabmIub6OJlW4wMbdW/rmb39XD59RjM7cxiRvqoYVFX12ZArPKXwfvYoL9DV5zuslx4+H2v3gz+5ake0zMret3eWU9gFUm5zATczUVbivas0xOof9aaWBlxuYzAWzt+vv7c2DA1e8AAABbg4i4OjPdDGqzuGMPAAAAAECLMbEHAAAAAKDFmNgDAAAAANBiTOwBAAAAAGixyWPdgcbWSVpQxOabvLLA2jUm5xkm5grXlUX2XBW57U1sfxN7momVhdiua5Aj+UJ8SwY3Zz7f5BxlYnvVQ/PXFwFzjAe7QnPOzSZ2RNGeYXJuNDF3/ncq2vuZnLLAniTtY2Ll9SXVrxWXs62JuX6UxQtd8bzyeP7M5AAAAABAF+7YAwAAAADQYkzsAQAAAABoMSb2AAAAAAC0GBN7AAAAAABarD3F8/aQ9OUGeWVRt7/fvZ6Td9ZjYd7juLuoUrerq95WVoKTpFtMbDvTj18Pbh9mNisL2UnSChO7qmgfbHJWmlhZbFCS1jR4PVeAsOyDJN1nYquLdllUTpL2NLF5JlYWBNxxb5P0qImZynX7mkqLZdqhrirhK03shyb2os20JelPBzenLzE5AAAAAPAY7tgDAAAAANBiTOwBAAAAAGgxJvYAAAAAALQYE3sAAAAAAFqsPcXzFkt6XxEri7xJUlHvTheaQnl9Zrs15YaSDixzHqjn3HZuPbaP2b+p1abpRftX9ZR119Vj7rA3FO1fmpydTGyxif130TYl5WqnWZL2N7EZJrbw7MFtV36wrIE41L4OL9q/1a21HFfKcC8TczX8Dina5+uyWs6fm9gNZl+36WeD2rfrw7Wc44u2G2sAAAAA6MYdewAAAAAAWoyJPQAA41REnBgRAxExsHTp0rHuDgAAGKeY2AMAME5l5umZ2Z+Z/X19bh0ZAABAm9bYz5J0XBHb1eQtL9rvMTlu4fJ9JrZL0Z5mcvYzsUkm5t5CmVy+QL1jU9bV1/VP+ZbZV3Ej5+WmHIDr/0Er6rE/KRfQLzD72tPE3P853SL+0rNMbIqJbW9i5w1uHuz64Oob3GFibmH/zoObf7WvyTFje6ipl3BoWSTgIbOvoq7D1M+bHAAAAADowh17AAAAAABajIk9AAAAAAAtxsQeAAAAAIAWY2IPAAAAAECLtad43naSDitiO+xuEh8t2k+vpzx8WT0201VFK6uu7W9yDjGxhSZmqqnV+rpbPWWK2deb3bCV1f/2MDmuitxBJlYW8bvJ5FxnYl82sUtM7I+K9nyTY86F2/+rf1gEXmG2c2Pk+v9uEzuiaM83Ocea2DkmdlfR3tnkzB3c/N6ZJgcAAAAAHsMdewAAAAAAWoyJPQAAAAAALcbEHgAAAACAFmNiDwAAAABAi7WneN59kk4tYtPurOdtLNqL76nnTDP7v/HmeqysuTb7inrOPZ+vx+aZ/a8wsUnl/k3OehNbYmIzivaC6+s5B7rtvmqChbKWniTdaGL7H1yPlccoSTsVbXdu1pqYqz+4uGjPPrees9Jst8rEbjGxE3+w+X1N+Ww9dofJ+5+iXfZdkj5QtN1YAwAAAEAX7tgDAAAAANBiTOwBAAAAAGgxJvYAAAAAALQYE3sAAAAAAFqsPcXzpkvqN7HSdUV7L5PjisEdY2LbFO0NJscVpHN5rqDe0qJdFsCT/Fsv5XZSvahbWfhPklab2DUmdkDRnmpynm5iF5jYtiY2v2gvMDllgT1JWmdiy02s5AreucJ1Lu/ior2fyXnAxO4ysX2L9vNMTnn9/t7kAAAAAEAX7tgDAAAAANBiTOwBAAAAAGgxJvYAAAAAALQYE3sAAAAAAFqsPcXz1klaVMRWmbyykNl6k+OK5+1iYn1Fe63J+amJzTIxVxiv5IriXWViZRE2t39X0M2di4dMrCxIVxYRlKRbTGz7Bv2SpNlF250vty93fsqihK6v95lYObaStMTEdi7a7po7zMTKQo9S/ThdX+8p2v9tcgAAAACgC3fsAQAAAABoMSb2AACMUxFxYkQMRMTA0qXusSUAAAAm9gAAjFuZeXpm9mdmf1+fW0MEAADQpjX2KyR9v0FeuW7ZrQUv12VL9bXNLs+tr77NxPY0sQUN9u/+zzbVxFxfy21vNjluTfeAiS0u2m6dvFvnv9DEVptYaaOJrTAx14/yvLraAm6M3Hl1N8PK15xmctx1OcnEyjFya/rLa8fVgwAAAACALtyxBwAAAACgxZjYAwAAAADQYkzsAQAAAABoMSb2AAAAAAC0WHuK520v6Zgi5oqulQXuDjM5K01sPxN7uGi7onVuu21NbJGJlQXcNpgcV+jPFX6bVbT3MDkPmdh6EysLuLmida6IXK8Fm12BOFeAcHaDbV0fdjYxVwjRFd5rwo2H60d5LbpihjcWbVeEDwAAAAC6cMceAAAAAIAWY2IPAAAAAECLMbEHAAAAAKDFmNgDAAAAANBi7Smet0H14m+ugFtZyGyJyXEF4242sbJonCu6t9TEykJ2Q21bFn5zOW5frohcWbBvrcnZxcRebGJlwb7pJuc2E3PFDN0YleM4xeS48+rOjytAV3L7dwUB3XGWY+SK2bnzuq+JlX11xRLL13PnFAAAAAC6cMceAAAAAIAWY2IPAAAAAECLMbEHAAAAAKDFmNgDAAAAANBi7Sme9yTVC7G5AmhlgbI+k+MKoJVFyxz3ek6TomhSvdCf65d7TZdXvqYruueKw+1sYmUhPscVh2tSINC9pttunYm5IngrinbTonhNx6h8TXc9uUJ/Lq/cl+vDbpvZBgAAAAAK3LEHAAAAAKDFmNgDAAAAANBiTOwBAAAAAGix9qyxn6z6uuVyjbpUf6tipslZZWLbm9jqBq/n1nQvNzG3rVv7XSrXkEv1OgJSszX2TdfAl6/ZpP6AJM0zsSZr2d16ene+nLLewPoGOU/EYhNz15N7y2xO0S5rRkj14+atNwAAAACbwbQBAAAAAIAWY2IPAMA4FREnRsRARAwsXeo+ggMAAICJPQAA41Zmnp6Z/ZnZ39fnPkcTAACAiT0AAAAAAK3WnuJ5GyQ90CBv26J9V8P9uwJorkhdaYqJuYJxTlnUzW3nisG5YnZlXxeZHPc2TpPCcq5fbjtXWM4pixI23b8rqFeeC3duXNHAjSbmlH1z10R5PEf5z7UAACAASURBVJIft/sa5JTXr8sBAAAAgC7csQcAAAAAoMWY2AMAAAAA0GJM7AEAAAAAaDEm9gAAAAAAtFh7iueF6oXFXAG0suhauc1QXF65/16Lz0nNC+qVpjXcV9kPV2iuaZG6XrdreowzinbTt5fca5bnp2kf3LXjti3zXPG86SbWpPBik77y1hsAAACAzWDaAAAAAABAizGxBwAAAACgxZjYAwAAAADQYkzsAQAAAABosfYUz3Pc2xK9Fqlruv/S+h63k6Q1PW7nlMftisM5Lq9JP5oUmhtqX2XsifR1S3LF+ZoUwWta4LDU5DxHgxwAAAAAExp37AEAAAAAaDEm9gAAAAAAtBgTewAAAAAAWqzda+ydKUXbrZF265/L7Vzelly/L9XX5zftV5NjcuvFyzX9rg9NrTWxJuvRJd+3kjtu1//yrampDV/P7avX8XXH3eQY3dtqK4t2Dr87AAAAACYW7tgDAAAAANBiTOwBAAAAAGgxJvYAAAAAALQYE3sAAAAAAFqsPcXzUr5gW2lj0V5qclyBNVc4zeWVVpmYK5w2zcTKvjquyJt7zZIritf0uMttmxbFc6/Z61tHrl/uXJSeSFE/19dyLN0xrm64/5I7nnKMosd9AwAAAJgwuGMPAAAAAECLMbEHAGCciogTI2IgIgaWLnWPoAEAADCxBwBg3MrM0zOzPzP7+/r6xro7AABgnGJiDwAAAABAi0VmjnUfGomIpZIWjXU/gFG2R2Zymw6AImKlpFvGuh8YZK6kZWPdCdQwLuMPYzI+MS7jzz6ZOaOXDVtTFZ/JDQBggrslM/vHuhN4TEQMMCbjD+My/jAm4xPjMv5ExECv2/IoPgAAAAAALcbEHgAAAACAFmNiDwBAO5w+1h1ADWMyPjEu4w9jMj4xLuNPz2PSmuJ5AAAAAACgjjv2AAAAAAC0GBN7AADGiYh4WUTcEhG3R8QHzPcjIk7tfP/6iDhwLPo50TQYl+M743F9RFwREfuPRT8nks2NSVfewRGxISJeM5r9m6iajEtEHBkR10bETRHx89Hu40TT4N+v7SPiBxFxXWdM3joW/ZxIIuKMiLg/Im4c4vs9/a5nYg8rIr4QER95nO//fUR8aTT7BABbs4iYJOnfJR0taV9Jb4iIfYu0oyXt1fk6UdLnR7WTE1DDcfmdpCMy89mSThHrVkdUwzHZlPdpSReObg8npibjEhGzJH1O0isz85mSXjvqHZ1AGv6svEfSzZm5v6QjJf1zREwZ1Y5OPGdJetnjfL+n3/VM7GFl5jsz8xTpf99Zvav4/icz88/HpncAsFV6jqTbM3NBZq6T9E1JxxY5x0r6SlaulDQrIp4y2h2dYDY7Lpl5RWY+2GleKWm3Ue7jRNPkZ0WS3ifpHEn3j2bnJrAm4/JGSd/NzDslKTMZm5HVZExS0oyICEnbSXpA0qOj282JJTMvU3Weh9LT73om9gAAjA+7Slrc1b6rExtuDras4Z7zt0u6YER7hM2OSUTsKuk4SV8YxX5NdE1+VvaWtENEXBoRV0fECaPWu4mpyZicJukPJN0j6QZJf5WZG0enexhCT7/rmdhvJSJiYUR8MCJujogHI+LMiJja+d5fdNZoPBAR50XELp14RMS/dtZ4PNRZw/GszvfOioiPR8R0Vf9B2SUiHul87RIRJ0fE1zq5P46I9xb9uS4iXtX5+zMi4qLO698SEa8bzXMDAC0RJlZ+dE2THGxZjc95RLxA1cT+pBHtEZqMyWcknZSZG0ahP6g0GZfJkg6S9ApJR0n6SETsPdIdm8CajMlRkq6VtIukAySdFhEzR7pjeFw9/a5nYr91OV7VD+fTVL0j+uGIeKGkf5T0OklPkbRI1WM4kvRSSc/v5M6S9KeSlnfvMDNXqVrncU9mbtf5uqd43W9IesOmRmftzh6Sfth5Y+CiTs6OnbzPRcQzt9RBA8BW4i5J87rau6m6gzLcHGxZjc55RDxb0pckHZuZy8vvY4tqMib9kr4ZEQslvUbV/z3+ZHS6N2E1/Tfsx5m5KjOXSbpMEsUmR06TMXmrquURmZm3q6oZ8oxR6h+8nn7XM7HfupyWmYsz8wFJn1A1iT5e0hmZeU1mrpX0QUnPjYj5ktZLmqHqhzcy8zeZeW8Pr3uupAMiYo9O+3hV/0CslXSMpIWZeWZmPpqZ16ha70Z1WgAY7CpJe0XEUzuFi14v6bwi5zxJJ3SeuDpU0kM9/ruN5jY7LhGxu6TvSnpzZt46Bn2caDY7Jpn51Mycn5nzJf2XpHdn5vdGv6sTSpN/w74v6fCImBwR0yQdIuk3o9zPiaTJmNwp6UWSFBE7SdpH0oJR7SVKPf2unzzy/cIo6l6LsUjVIzW7SLpmUzAzH4mI5ZJ2zcyfRsRpqqpl7h4R50r628x8eDgvmpkrI+KHqv6x+HTnzxM7395D0iERsaJrk8mSvjq8QwOArVtmPtpZ1nShpEmq3pS9KSLe2fn+FyT9SNLLJd0uabWqOy0YQQ3H5aOS5qi6KyxJj2Zm/1j1eWvXcEwwypqMS2b+JiJ+LOl6SRslfSkz7Ud+4Ylr+LNyiqSzIuIGVY+An9R5mgIjJCLOVvUJBHM7Bco/Jmkb6Yn9ro9MluZtDTqPmn1q0y+ziDhaVTGMSyUtz8z/24lPl7RC0l6ZubBr+x0lfVvS5Zn5kYg4S9JdmfnhiDhC0tczc7eu/JMlPT0z39RpH6fqonyXqjvyu2Xmxoh4g6S3ZeZLRvDwAQAAAGDC4lH8rct7ImK3iJgt6e8lfUvV2va3RsQBEbGtpE9K+lVmLoyIgyPikIjYRtIqSWskuSIz90maExHbP85r/0jV3fl/kPStrmqa50vaOyLeHBHbdL4Ojog/2CJHDAAAAAATHBP7rcs3JP1E1bqYBZI+npmXSPqIqrvo96oqrPf6Tv5MSf8h6UFVj+4vl/T/lTvNzN9KOlvSgohYsamqfpGzVtX6whd3+rEpvlJVkb7Xqyr6sETV4/rbPvHDBQAAAADwKP5WovMo/p9n5sVj3RcAAAAAwOjhjj0AAAAAAC3GxB4AAAAAgBbjUXwAAAAAAFqMO/YAAAAAALQYE3sAAAAAAFps8nCS506NnD+jCJZP8rsn+5vGRluM8P57PUb3dovra6/935LjMdL76lWv/Xoi12+vPwtNNN1uCx731Ru0LDP7Gr4ygBE2d+7cnD9//lh3AwAAjJCrr7665/9/D2tiP3+GNHBsEVxftDeYDcucofKa2GhivT53MKnH7Zpyx+j6X5pqYts0jDXRdIya9NXtq9fjbnpNNMlzr9fkOnwi12+Z57ZzMafcf9Ptej1Gc77iAS1q+KoARsH8+fM1MDAw1t0AAAAjJCJ6/v83j+IDAAAAANBiTOwBABinIuLEiBiIiIGlS5eOdXcAAMA4xcQeAIBxKjNPz8z+zOzv66PkBQAA8Ia1xl4zJR1VxF77syJwhtnQLRV4xMQOMrGFg5t5UT1lhdlshyn12P3rBrfLQoCSdIuJrTGxQ2cPbv/6gXrOfWa7p5vY6qJ90MUm6Z9NbGcTe3/R3s3kXGJi55jYhUV7lsn5nIktMbHvF213ItwxPsvEPlm0zZrTe0+ux57yKbOvHxXtPUzO103sgHropmsGt59prsE71tVj08zuFxftPU3O7Sa23MTKWhLuenbr7l9jYgAAAADGHe7YAwAAAADQYkzsAQAAAABoMSb2AAAAAAC0GBN7AAAAAABaLDKzcXL/lMiBnQbHDr9rcPsws91UE3vYxBaYWFkzbJ7JudnEXH2wspSZqzNmyp1pexO7oGjPNjkvbNivE4r2z03OgSbmStn9XdHez+T8oYm587pX0Xal+v7NxNx4Nzn3h5tYWb5Pkk4r2keYnFNM7AMmVh63uwaPNjFTrk/7Fm1XQtCNx20mtn/RXm9y3Lty7vota0S6mpHOS6WrM7O/YTqAEdbf358DA+5fHwAAsDWIiJ7//80dewAAAAAAWoyJPQAAAAAALcbEHgAAAACAFps8rOynSDppcOjyd59fJP3KbHiXie1mYjea2KLBzZuuqaf0mc12NO9ZXL9xcHul2c7FbjCxcsH5FSbnHhPbxsTmD27+0WtMzt0mZhbZXz6nCLi+uwXbbvH/hqJdLiKX9PcvMkG3yv7OYtx2MZuVhQskvXCfeuzTex9XRExVgu88UAtd9Nq9zYuWB/5oPWWdGUhXJOC6ou0W/7uCEG4xfvmS7lq6xcRWmVj58+EKKqwwsc+YGAAAAIBxhzv2AAAAAAC0GBN7AAAAAABajIk9AAAAAAAtxsQeAAAAAIAWG17xvPsknVrEzjxmcNsVfpthYueaWFmsTarXNnOF39x2e26sx4o6fK743P0X12OmFJxWF21XE217E9vLxKYX7UkLTZKpGai19dC1RZG1A15Zz9lwXj02aTuz/7Kz+5mcz5sjX2DyyuFwxfpmm5gb27XFxeMK8bn9/+OttdDD/zO4PfP5ZruyUKLkr+myr99vuK+dTKx0mIm5Qnnu56Ps6yEN90XxPAAAAKAVuGMPAAAAAECLMbEHAAAAAKDFmNgDAAAAANBiTOwBAAAAAGixYRXPe2it9KNbBsfOL3L6B+rb7Wv2tY2JLTaxZxTtJSZnqYmt/J967IGivc5s54rbucJ4Zd09V1fO1ZAztfl0YNF+8zn1HFeTsM/Efla0bzCF8lxNtPseqcdmF+ewz5zTn5t9OeUYudpzx5qYuybK2GHX1XNesm09doYpNlie+x9dVs95k+mDG9sji9f8qXm9p5vtXJ2/vacNbi8vqzVKmrNHPXZnWSBS9fO1q3k9V6MQAAAAQDtwxx4AAAAAgBZjYg8AwDgVESdGxEBEDCxd6p5PAwAAYGIPAMC4lZmnZ2Z/Zvb39bkFWAAAAEzsAQAAAABotcjMxsn90yMHnlUEP1u0y2pkQ3HVx1zlutKKhvvf4TgT/EXRNqXMHryiHnOV3ia/tgiYMnJ33l+PzTb7Orto/5nJOcnEnElF+1Mmx1W8c1Xqykp8603Oq01sjomV59BVVHQV3LY72gTPKNruoniHiR1SD936T4Pbe5fjKknfN7Ej6qH7Lxrc3nHPes4yc+HPNLv/adE2hfJ0l4ndbWLleLjKlWVlSUlxvK7OzH6TDWAM9Pf358CAqVALAAC2ChHR8/+/uWMPAAAAAECLMbEHAAAAAKDFmNgDAAAAANBiw1pjv1dE/lsRu6poH2y2MyvZtbd5S+HKjfXYodsNbt/6SD3HLA/WoW4d8bSiPcvkOPNMbEmDnDUm5taRF0smf2vOg/uQI1cfuVw+X3ZTkl5lYm7Je7lyfY4Zs4tMX92+dt22CLjOH2ZiK+uhdRcMbk/Zrp6jl9dDt367Htt7l8HtO+6p55iV8rXyA1L9uMtl8pJ0lIldbmLzi7YrSeHKNUw1sdrPgluvb85z3Mcae2A8YY09AABbN9bYAwAAAAAwQTGxBwAAAACgxZjYAwAAAADQYkzsAQAAAABoscnDSV4j6eYiVhYMc8W7XJ00TaqHpppCbCpic0zKQyZ26/rNJ240Gz7DVfq7xsTKt0TcQZo+pKl7VBb/W2V25QqluRpoZczV79vFxB42sbLWoCs2eISpXFheI5K0dG2x77vqObNMcbsdTWG8KXMHt69eVs+ZZPa1s+nXwqJYnjvPpp6eXrhTPbbhvsHtN+5jNjRV8Oab6+SWor3O7Ko2PpJmmFiNuSgevq8eAwAAANAO3LEHAAAAAKDFmNgDAAAAANBiTOwBAAAAAGgxJvYAAAAAALRYZGbj5P7tIwcOLYL/t2i76nllJTBJWmFiprBYrdLb80yOK7p3oIktLdobTM4iE3PKqoHLTc6NJuZe89yi/X8a5EjSXiZWVk871uS46na/MrHymFxlNhdz41heF+6acOfmxSY2vWibAn66zcSc8jo8yuT83MRcYbxS03Pj9l/mHWFyVpuYq5bYX7TddWP6FZ/S1ZlZbg1gjPT39+fAgKnACgAAtgoR0fP/v7ljDwAAAABAizGxBwAAAACgxZjYAwAAAADQYkzsAQAAAABoscnDSb73YemTPxkcu7hoH2y2K+vfSVKfif2rie1ftF3dr9km5mqU7Vy0Xf227U3sdhPbo2i7mmtXmZgzr2j/3BTKO8Bs94cmVtZhc7UAv2BirtbclKI93wzk3ffUY7tuV4+temRwe6F5vfI8SNLM8+qxJesHt13NvV3n1mPfWlaPlTUQ13y9nrON2f+PTOxdRdvVTtzTxM40sVll4Jx6ziFmu8tNTD8b3NzPpLhzDwAAAKAduGMPAAAAAECLMbEHAGCciogTI2IgIgaWLi0/sxUAAKDCxB4AgHEqM0/PzP7M7O/rc4vYAAAApMjMxsnPjMhvl7Fysbm7oVDmSNJD9dBFZr12+d+YfcyupprYgImVa7EnmZxVJrbexF4yrQgca5LcIntXcKBczP5/TM7PTMwdeLnI2hUlqC3glrTcxOYU7SNMzkoTu9vEyuuiLHjgcoZ6zXJh/DUm5wITc4vSzy7aR5ucJQ36INUvKHfdu+N2F9jFJlY6xsTK61KSzi/aq02OOcb4lK7OzP4GPQEwCvr7+3NgwP12AwAAW4OI6Pn/39yxBwAAAACgxZjYAwAAAADQYkzsAQAAAABoMSb2AAAAAAC02LCK5/UfFDnwyyK4psGGF5mYKZ5XK/IlSc8q2n9mclxBuieb9yzu3Ti4/ZTZ9Zw7HqjHnvYC8wIfKtovMjkXmtgKE/ti0X6FyXHV7Sab2JVF21XK287E3EA+WrT3r6cs+3A9dp/ZVVkFcceZ5uUerscm71mP/X7B4HZZFVHyBeIWmljZV1d87mATcwX19i4r711tkt5hYq6M49qi7Qpm/YOJuevky0XbDNAjV9RCMYPiecB4QvE8AAC2bhTPAwAAAABggmJiDwAAAABAizGxBwAAAACgxZjYAwAAAADQYsMrnrdb5MD7iuDZRdsVMnP14g6sh+4/rx7bcaciYOq3abqJrTOxRUV7qskpX0+S9jOxQ4r2C03ObSZ2sYnNKNrnmhxT588WDSxr/7nxeJeJ3Wxi1xTtw03OVSbm6vVtU7TdmLnz5Y6x3P89Jscdj6sPeFjRdteqq0f3cxN7W9Euj3momKnXWCvO547xL03sBhO7rmi7c//beiguo3geMJ5QPA8AgK0bxfMAAAAAAJigmNgDAAAAANBiTOwBAAAAAGgxJvYAAAAAALTY5GFlr1a9oNori7YrUOaK1M2rh3Z0hfFWFm1X+MsVPFtlYmWRsp1NzkITc/0vi9KV50WSFpvYWhMrC6q9weRcaGJlMUCpfr6e13Bfrv9l0TX3eu6toSkmNq1oLzU57ny5Mdqzwb5+amLzTawsgvd0k+NqVZXnWZK+W7TNNW6PxxXUK8+FyynHR/LX6sFF2xWWdD9DAAAAAFqBO/YAAAAAALQYE3sAAAAAAFqMiT0AAAAAAC3GxB4AAAAAgBYbXvG8UL2IV1nkyxWHW25i95iYK7w3qWivNzm7mJjLm1O0Z5gcV3zMFVR7qGjfZ3JcgTVXBK08P/9qcg4zMVcEbWHRPtXklMXnpHoxQEl6RtF2BQl/aWLuGPcp2q5Y23wTu6Ieevhnm3+5J29ntruxHptZFt67yuzMXCd3LKvH9ixi9/xPPWfXp5r9u5+ZkvvZ2NgwrzSrQQ4AAACA1uCOPQAAAAAALcbEHgCAcSoiToyIgYgYWLrUfbYnAAAAE3sAAMatzDw9M/szs7+vr2+suwMAAMap4a2xf0Da8PXBoXJ5uFs6PdPEZpvYEhMrl8VPd29FzDMxt1a+WH/8+0X1lHLpvOTXcJfLzd1yZ9dVl+fORWnmggZJkpavHtyeM80kufoGbv18sS97Ig4xMZdXjsf+JqespyBJ5v+xM3cuAneb7X5ltvtjk1desO66MQP5tINNXjFGu+5kcgZMzCn69bC5cGb+1GxnakvcWWxb/kxJ0nTXVwAAAACtwB17AAAAAABajIk9AAAAAAAtxsQeAAAAAIAWY2IPAAAAAECLDa943jbSpKKY2Y77FG1XrG2Kia2sh3bfYPLK/bniZtNNbBcTK4rGPbms/CfpyabommbUQ3PKfphCfJplYq5SXpm3r8lxVQnNOZxT9rUsNCfVighKkhabWFlQzx3PfBNz10C5f1dRcVsT28/Eyv6743mhiblrYk8TK7lC1O41y/27HDe27hOsinGc6a57d02YInu7u3Fr4tQetwMAAAAwqrhjDwAAAABAizGxBwAAAACgxZjYAwAAAADQYkzsAQAAAABoseEVz+uT9K4iVha8cwXwnNUN88pic64gmSvO55TFzfYyOYeamHv7o+y/KcRni/q5gmfzivZtJscVwXOF3x4o2ktMjuMK3pVF11xOWWBPalbUzZ0vd4yu/2UhPlNE0O5rgYmVxQZdITtXPG+SiZXnYn3D7dw1XV4795gcd9xN3qozxSBtvwAAAAC0AnfsAQAAAABoMSb2AAAAAAC0GBN7AAAAAABajIk9AAAAAAAtNrzieeskLSxiZZEvVwhsexNzeVNNzBUzK7kiYk5Z8M4Vg3OFxbYxsbIYXJNiapIvBlfu3+3rbhNzfS3PV9OCdK6YYVm40BWRc28NuUJv5TGW50/y58udi7LYoBv/soig5M9FWeyx3Lfkr9Umb4m568tt54o/luPofjbcden6Wp5X1weK5wEAAACtxR17AAAAAABajIk9AAAAAAAtxsQeAAAAAIAWG94a+zWSbili5Tpvt+57nYmtNTG3ZrhJjltj7dYkl+up15sct6bf7ausLVDueyizTKxc3+765dZAN6kt4NZcuz641yy3dWvgXR9crHwLya3pd31dYWIld27cddJkDb8bRzf+bl9lnuv7zB775Y6naaz8mXTH48YfAAAAQCtwxx4AAAAAgBZjYg8AAAAAQIsxsQcAYJyKiBMjYiAiBpYubfL5rwAAYCJiYg8AwDiVmadnZn9m9vf19Y11dwAAwDg1vOJ5UyXtU8RcEbTSRhNzRb5cAa8mbz1syUJpZVG8ofLKgoCugJuLNdm/KyLnuPNaxtz4uONxfS339YDJcePjCiiW+3fnwXH7KsfbnQfXL5dXcufGXZdTGuzfvZ4bW3fuy34sNzmOe83yGnDnxhXUAwAAANAK3LEHAAAAAKDFmNgDAAAAANBiTOwBAAAAAGgxJvYAAAAAALTY8IrnheqFy2YXbVcIzHFFylys1KTQmOSLs23JtzGaFOJzMVforzSrYR96PV9l4b+h8sr9Ny0Q2ETT4nYur8lrNj3uLansf6/jvyX70FSTawkAAADAuMQdewAAAAAAWoyJPQAAAAAALcbEHgAAAACAFmNiDwAAAABAiw2veJ5TFt1a3yBH8sXNmrzN8ESKfDUpLOb64GK99sOdnybH7frepIhc03PvbMn9l8fYtMiby2uy7RoT67XQX9OxLvfvXs+Nf6+v2evxjPS+AAAAAIwq7tgDAAAAANBiTOwBAAAAAGgxJvYAAAAAALTY8NbYh6QpRWybot10DfG2JubW+ZZrjV2OW0/tNF3XXWry9kfTfk01sV7X/rvtmoyHW7/tYuW25b6H2r/ra5P1503rLpTHvc7kuL72Wm/AjZkb22kNcpqew+mb65Savy1XvubahtsBAAAAaAXu2AMAAAAA0GJM7AEAAAAAaDEm9gAAAAAAtBgTewAAAAAAWmx4xfM2ql4Q7OGivdps5wqGNSkOJtULnjUtSOeUr1n2fah9ueJmZaxpH1wBt1JZoHAo7rzOaJDj+ur6VY6l25cr4Leqwf7dMboieK6v5bbumnAxd4xNCi+6PpTn2eW511tpYu68uljJnXt3rZbj6ManSRFBAAAAAOMSd+wBAAAAAGgxJvYAAIxTEXFiRAxExMDSpUvHujsAAGCcYmIPAMA4lZmnZ2Z/Zvb39fWNdXcAAMA4xcQeAAAAAIAWi8xsnhyxVNKikesOgHFkj8zkFiEwTkTESkm3jHU/MMhcScvGuhOoYVzGH8ZkfGJcxp99MtOV6d6sYVXF5z/5AACMmVsys3+sO4HHRMQAYzL+MC7jD2MyPjEu409EDPS6LY/iAwAAAADQYkzsAQAAAABoMSb2AAC0w+lj3QHUMCbjE+My/jAm4xPjMv70PCbDKp4HAAAAAADGF+7YAwAAAADQYkzsAQAYJyLiZRFxS0TcHhEfMN+PiDi18/3rI+LAsejnRNNgXI7vjMf1EXFFROw/Fv2cSDY3Jl15B0fEhoh4zWj2b6JqMi4RcWREXBsRN0XEz0e7jxNNg3+/to+IH0TEdZ0xeetY9HMiiYgzIuL+iLhxiO/39LueiT0AAONAREyS9O+Sjpa0r6Q3RMS+RdrRkvbqfJ0o6fOj2skJqOG4/E7SEZn5bEmniHWrI6rhmGzK+7SkC0e3hxNTk3GJiFmSPifplZn5TEmvHfWOTiANf1beI+nmzNxf0pGS/jkipoxqRyeesyS97HG+39Pveib2AACMD8+RdHtmLsjMdZK+KenYIudYSV/JypWSZkXEU0a7oxPMZsclM6/IzAc7zSsl7TbKfZxomvysSNL7JJ0j6f7R7NwE1mRc3ijpu5l5pyRlJmMzspqMSUqaEREhaTtJD0h6dHS7ObFk5mWqzvNQevpdz8QeAIDxYVdJi7vad3Viw83BljXcc/52SReMaI+w2TGJiF0lHSfpC6PYr4muyc/K3pJ2iIhLI+LqiDhh1Ho3MTUZk9Mk/YGkeyTdIOmvMnPj6HQPQ+jpd/3kEesOAAAYjjCx8qNrmuRgy2p8ziPiBaom9s8b0R6hyZh8RtJJmbmhuhGJUdBkXCZLOkjSiyQ9WdIvI+LKzLx1pDs3QTUZk6MkXSvphZKeJumiiLg8Mx8e6c5hSD39rmdiDwDA+HCXpHld7d1U3UEZbg62rEbnPCKeLelLko7OzOWj1LeJqsmY9Ev6tR7wAgAAAZFJREFUZmdSP1fSyyPi0cz83uh0cUJq+m/YssxcJWlVRFwmaX9JTOxHRpMxeaukT2X1Gei3R8TvJD1D0q9Hp4swevpdz6P4AACMD1dJ2isintopXPR6SecVOedJOqFTMfdQSQ9l5r2j3dEJZrPjEhG7S/qupDdz53FUbHZMMvOpmTk/M+dL+i9J72ZSP+Ka/Bv2fUmHR8TkiJgm6RBJvxnlfk4kTcbkTlVPUCgidpK0j6QFo9pLlHr6Xc8dewAAxoHMfDQi3quqgvckSWdk5k0R8c7O978g6UeSXi7pdkmrVd1pwQhqOC4flTRH0uc6d4gfzcz+serz1q7hmGCUNRmXzPxNRPxY0vWSNkr6Umbaj/zCE9fwZ+UUSWdFxA2qHgE/KTOXjVmnJ4CIOFvVJxDMjYi7JH1M0jbSE/tdH9VTFwAAAAAAoI14FB8AAAAAgBZjYg8AAAAAQIsxsQcAAAAAoMWY2AMAAAAA0GJM7AEAAAAAaDEm9gAAAAAAtBgTewAAAAAAWoyJPQAAAAAALfb/AwXxtg6ORPs7AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1440x360 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot mfcc \n",
    "\n",
    "plot_mfccs(mfccs)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}