Molecular_Variational_Autoencoder.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "37706c9c-1ef4-4051-b17e-e54aa0664180",
   "metadata": {},
   "source": [
    "# Variational Autoencoder for Molecular graph images\n",
    "\n",
    "This block\n",
    "\n",
    "- imports all modules/libraries needed\n",
    "- contains a function for a learning rate scheduler (if needed)\n",
    "- reads in SMILES strings and creates a list\n",
    "- featurizes via the SmilesTo Image featurizer from Deepchem (https://github.com/deepchem/deepchem)\n",
    "- Four channel image used for featurization\n",
    "  * Atomic numbers + bonds\n",
    "  * partial charges\n",
    "  * valence\n",
    "  * hybridization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "979d0bc1-0585-4b38-903d-b64edabe0891",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\ProgramData\\Anaconda3\\envs\\rdkitenv\\lib\\site-packages\\keras\\src\\losses.py:2976: The name tf.losses.sparse_softmax_cross_entropy is deprecated. Please use tf.compat.v1.losses.sparse_softmax_cross_entropy instead.\n",
      "\n",
      "WARNING:tensorflow:From C:\\ProgramData\\Anaconda3\\envs\\rdkitenv\\lib\\site-packages\\tensorflow\\python\\util\\deprecation.py:588: calling function (from tensorflow.python.eager.polymorphic_function.polymorphic_function) with experimental_relax_shapes is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "experimental_relax_shapes is deprecated, use reduce_retracing instead\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Skipped loading modules with pytorch-geometric dependency, missing a dependency. No module named 'torch_geometric'\n",
      "Skipped loading modules with pytorch-geometric dependency, missing a dependency. cannot import name 'DMPNN' from 'deepchem.models.torch_models' (C:\\ProgramData\\Anaconda3\\envs\\rdkitenv\\lib\\site-packages\\deepchem\\models\\torch_models\\__init__.py)\n",
      "Skipped loading some Jax models, missing a dependency. No module named 'jax'\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of features and datapoints:  (6731, 80, 80, 4)\n",
      "featurization done with:  SmilesToImage\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import pandas as pd \n",
    "import deepchem as dc\n",
    "import time\n",
    "from rdkit import Chem\n",
    "from functools import partial\n",
    "import matplotlib.pyplot as plt\n",
    "from rdkit.Chem import AllChem, Draw\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "def exp_decay(lr0,s):\n",
    "    def exp_decay_fcn(epoch):\n",
    "        return lr0*0.1**(epoch/s)\n",
    "    return exp_decay_fcn\n",
    "\n",
    "#df = pd.read_csv(\"392-azo.csv\")\n",
    "#df = pd.read_csv(\"212-azo.csv\")\n",
    "#df = pd.read_csv(\"604-azo.csv\")\n",
    "#df = pd.read_csv(\"621-azo.csv\")\n",
    "df = pd.read_csv(\"6731-azo.csv\")\n",
    "\n",
    "Xa = [\"\"]*len(df)\n",
    "w = [\"\"]*len(df)\n",
    "i=0\n",
    "for name in df[\"SMILES\"]:\n",
    "    Xa[i]=name\n",
    "    w[i]=1.0\n",
    "    i += 1\n",
    "Xa = list(Xa)\n",
    "w=list(w)\n",
    "\n",
    "#===========================================================================================\n",
    "#featurizer\n",
    "\n",
    "featurizer=dc.feat.SmilesToImage(img_spec=\"engd\")\n",
    "featname=\"SmilesToImage\"\n",
    "\n",
    "f = featurizer.featurize(Xa)\n",
    "print(\"Number of features and datapoints: \",f.shape)\n",
    "\n",
    "#===========================================================================================\n",
    "print(\"featurization done with: \",featname)\n",
    "\n",
    "# set npp for how many points in the dataset, len(df) for whole dataset\n",
    "nnp=len(df)\n",
    "\n",
    "#set y to lambdamax\n",
    "y = df[\"Lmax\"]\n",
    "\n",
    "#set dataset and split\n",
    "\n",
    "dataset = dc.data.NumpyDataset(X=f[:nnp,:,:,:], y=y[:nnp], ids=Xa[:nnp], w=w[:nnp])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "40fa585e-c87d-4f2d-ad70-31c6fab66be1",
   "metadata": {},
   "source": [
    "## Plot original image versus image reconstructed with the model\n",
    "\n",
    "- new_images function takes an image from the featurizer and runs it through the autoencoder\n",
    "- resulting image is the reconstructed image from the model, showing the selected channel\n",
    "- old_images takes the image from the featurizer and shows the selected channel\n",
    "- options for rotation and cropping are also available in the above two functions\n",
    "- plots original image on top, channel by channel, and reconstructed image on the bottom, channel by channel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "9c7efe7f-b2a8-4643-a8c1-5216b69b6360",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 316ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "1/1 [==============================] - 0s 67ms/step\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "1/1 [==============================] - 0s 43ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x1200 with 16 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "from scipy import ndimage\n",
    "\n",
    "#load a model\n",
    "#autoencoder = tf.keras.models.load_model(\"6731_30March_test_autoencoder\")\n",
    "\n",
    "\n",
    "cmap=\"tab20c\"\n",
    "\n",
    "def new_images(model,image,channel):\n",
    "    recons = model.predict(image)[0,:,:,channel] #last number cycles through channels 0-3\n",
    "    rot = ndimage.rotate(recons, 0)\n",
    "    rot = rot[20:60, 20:60]\n",
    "    return rot\n",
    "\n",
    "def old_images(model,image,channel):\n",
    "    recons = image[0,:,:,channel] #last number cycles through channels 0-3\n",
    "    rot = ndimage.rotate(recons, 0)\n",
    "    rot  = rot[20:60, 20:60]\n",
    "    return rot\n",
    "    \n",
    "m1=featurizer.featurize(\"[H]C1=CC([N+]([O-])=O)=CC([N+]([O-])=O)=C1/N=N/C2=CC([H])=C(C=C2[H])N(CC)CC\")\n",
    "m2=featurizer.featurize(\"CC1=C(/N=N/C2=CC=C(N(C)C)C=C2C)C=CC(C)=C1\")\n",
    "\n",
    "newimages=[]\n",
    "oldimages=[]\n",
    "newimages2=[]\n",
    "oldimages2=[]\n",
    "for i in range(4):\n",
    "    new1 = new_images(model=autoencoder,image=m1,channel=i)\n",
    "    old1 = old_images(model=autoencoder,image=m1,channel=i)\n",
    "    newimages.append(new1)\n",
    "    oldimages.append(old1)\n",
    "    new1 = new_images(model=autoencoder,image=m2,channel=i)\n",
    "    old1 = old_images(model=autoencoder,image=m2,channel=i)\n",
    "    newimages2.append(new1)\n",
    "    oldimages2.append(old1)\n",
    "\n",
    "#can insert scaling code here\n",
    "\n",
    "fig, arrax = plt.subplots(4, 4, figsize=(12,12))\n",
    "for i in range(4):\n",
    "    arrax[0,i].imshow(oldimages[i],cmap=cmap)\n",
    "    arrax[1,i].imshow(newimages[i],cmap=cmap)\n",
    "    arrax[2,i].imshow(oldimages2[i],cmap=cmap)\n",
    "    arrax[3,i].imshow(newimages2[i],cmap=cmap)\n",
    "    arrax[0,i].axis(\"off\")\n",
    "    arrax[1,i].axis(\"off\")\n",
    "    arrax[2,i].axis(\"off\")\n",
    "    arrax[3,i].axis(\"off\")\n",
    "arrax[0,0].set_title(\"Original Images (channels 1-4)\")\n",
    "arrax[1,0].set_title(\"Reconstructed Images (channels 1-4)\")\n",
    "arrax[2,0].set_title(\"Original Images (channels 1-4)\")\n",
    "arrax[3,0].set_title(\"Reconstructed Images (channels 1-4)\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f0dc88c3-17a9-48af-8032-4b0cf8d34951",
   "metadata": {},
   "source": [
    "## Define the sampling layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "0a3fa376-fceb-4b59-8ef1-0ae20a1718c1",
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import backend as K\n",
    "\n",
    "class Sampling(tf.keras.layers.Layer):\n",
    "    def call(self,inputs):\n",
    "        z_mean, z_log_var = inputs\n",
    "        batch = tf.shape(z_mean)[0]\n",
    "        dim = tf.shape(z_mean)[1]\n",
    "        epsilon = K.random_normal(shape=(batch,dim))\n",
    "        return z_mean + tf.exp(0.5*z_log_var)*epsilon"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7aae5b43-eed5-4e85-b0c0-c2693688b2c5",
   "metadata": {},
   "source": [
    "## Variational Autoencoder\n",
    "\n",
    "- uses Conv2D and MaxPooling in the encoder, Conv2DTranspose for the decoder\n",
    "- 4 channel images from SmilesToImage from DeepChem\n",
    "- regularization (l2) included and padding=\"same\"\n",
    "- borrows architechture from https://github.com/davidADSP/Generative_Deep_Learning_2nd_Edition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "b4be5e51-b8d8-4d00-bc94-252fbe1b4bcb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\ProgramData\\Anaconda3\\envs\\rdkitenv\\lib\\site-packages\\keras\\src\\engine\\base_layer_utils.py:384: The name tf.executing_eagerly_outside_functions is deprecated. Please use tf.compat.v1.executing_eagerly_outside_functions instead.\n",
      "\n",
      "INFO:tensorflow:Assets written to: ConvAE\\ConvAE\\assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: ConvAE\\ConvAE\\assets\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "190/190 - 30s - loss: 0.1881 - mse: 0.1828 - val_loss: 0.1847 - val_mse: 0.1848 - 30s/epoch - 157ms/step\n"
     ]
    }
   ],
   "source": [
    "alphac = 0.0\n",
    "embsize=128\n",
    "epcs = 1\n",
    "\n",
    "#=======================================================================================================\n",
    "encoder_input = tf.keras.layers.Input(shape=[80,80,4], name = \"encoder_input\")\n",
    "x = tf.keras.layers.Conv2D(filters=32,kernel_size=(3,3), strides=2,padding=\"same\",\n",
    "                           kernel_regularizer=tf.keras.regularizers.l2(alphac))(encoder_input)\n",
    "x = tf.keras.layers.BatchNormalization()(x)\n",
    "x = tf.keras.layers.ReLU()(x)\n",
    "x = tf.keras.layers.Conv2D(filters=64,kernel_size=(3,3), strides=2, padding=\"same\",\n",
    "                          kernel_regularizer=tf.keras.regularizers.l2(alphac))(x)\n",
    "x = tf.keras.layers.BatchNormalization()(x)\n",
    "x = tf.keras.layers.ReLU()(x)\n",
    "x = tf.keras.layers.Conv2D(filters=128,kernel_size=(3,3), strides=2, padding=\"same\",\n",
    "                          kernel_regularizer=tf.keras.regularizers.l2(alphac))(x)\n",
    "x = tf.keras.layers.BatchNormalization()(x)\n",
    "x = tf.keras.layers.ReLU()(x)\n",
    "shape_before_flattening = K.int_shape(x)[1:]\n",
    "x = tf.keras.layers.Flatten()(x)\n",
    "\n",
    "z_mean = tf.keras.layers.Dense(embsize,name = \"z_mean\")(x)\n",
    "z_log_var = tf.keras.layers.Dense(embsize,name = \"z_log_var\")(x)\n",
    "z = Sampling()([z_mean,z_log_var])\n",
    "\n",
    "encoder = tf.keras.models.Model(encoder_input,[z_mean,z_log_var,z], name = \"encoder\")\n",
    "#=======================================================================================================\n",
    "decoder_input = tf.keras.layers.Input(shape=(embsize,), name = \"decoder_input\")\n",
    "x = tf.keras.layers.Dense(np.prod(shape_before_flattening))(decoder_input)\n",
    "x = tf.keras.layers.BatchNormalization()(x)\n",
    "x = tf.keras.layers.ReLU()(x)\n",
    "x = tf.keras.layers.Reshape(shape_before_flattening)(x)\n",
    "x = tf.keras.layers.Conv2DTranspose(filters=128,kernel_size=(3,3),strides=2, padding=\"same\",\n",
    "                          kernel_regularizer=tf.keras.regularizers.l2(alphac))(x)\n",
    "x = tf.keras.layers.BatchNormalization()(x)\n",
    "x = tf.keras.layers.ReLU()(x)\n",
    "x = tf.keras.layers.Conv2DTranspose(filters=64,kernel_size=(3,3),strides=2, padding=\"same\",\n",
    "                          kernel_regularizer=tf.keras.regularizers.l2(alphac))(x)\n",
    "x = tf.keras.layers.BatchNormalization()(x)\n",
    "x = tf.keras.layers.ReLU()(x)\n",
    "x = tf.keras.layers.Conv2DTranspose(filters=32,kernel_size=(3,3),strides=2, padding=\"same\",\n",
    "                          kernel_regularizer=tf.keras.regularizers.l2(alphac))(x)\n",
    "x = tf.keras.layers.BatchNormalization()(x)\n",
    "x = tf.keras.layers.ReLU()(x)\n",
    "decoder_output = tf.keras.layers.Conv2D(4,kernel_size=(3,3),activation=\"sigmoid\",padding=\"same\",name=\"decoder_output\")(x)\n",
    "\n",
    "decoder = tf.keras.models.Model(decoder_input,decoder_output)\n",
    "#=======================================================================================================\n",
    "outputs = decoder(z)\n",
    "autoencoder = tf.keras.models.Model(encoder_input,outputs)\n",
    "\n",
    "# create loss for reconstruction plus the loss for the PDF for the embeddings ==========================\n",
    "\n",
    "scale_ll = 1.0 / 16000.0\n",
    "latent_loss_arg = -0.5 * tf.reduce_sum(1+ z_log_var - tf.exp(z_log_var)-tf.square(z_mean),axis=-1)\n",
    "latent_loss = scale_ll * tf.reduce_mean(latent_loss_arg)\n",
    "reconstruction_loss = tf.keras.losses.mean_squared_error(encoder_input,outputs)\n",
    "vae_loss = K.mean(reconstruction_loss + latent_loss)\n",
    "autoencoder.add_loss(vae_loss)\n",
    "\n",
    "optimizer = tf.keras.optimizers.SGD(learning_rate = 0.001, momentum = 0.9)\n",
    "checkpoint_cb = tf.keras.callbacks.ModelCheckpoint(r\"ConvAE/ConvAE\",monitor=\"val_mse\",save_best_only=True)\n",
    "autoencoder.compile(optimizer=optimizer,metrics=[\"mse\"])\n",
    "\n",
    "history = autoencoder.fit(dataset.X,dataset.X,epochs=epcs,verbose=2,validation_split=0.1,callbacks=[checkpoint_cb])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "12d6448c-8cef-40ba-b0a7-db1250a148d1",
   "metadata": {},
   "source": [
    "## Generate new images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "67b11efd-d5fa-4ebf-b03b-94ce46795f2e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 197ms/step\n",
      "(2, 80, 80, 4)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 8 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#load a model\n",
    "#decoder = tf.keras.models.load_model(\"6731_30March_test_decoder\")\n",
    "#embsize=256\n",
    "\n",
    "z_gen = np.random.normal(size=(2,embsize))\n",
    "\n",
    "recons_embed = decoder.predict(z_gen)\n",
    "print(recons_embed.shape)\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "from scipy import ndimage\n",
    "\n",
    "cmap=\"tab20c\"\n",
    "\n",
    "def new_images(image,channel):\n",
    "    recons = image[:,:,channel] #last number cycles through channels 0-3\n",
    "    rot = ndimage.rotate(recons, 0)\n",
    "    rot  = rot[20:60, 20:60]\n",
    "    return rot\n",
    "\n",
    "newimages=[]\n",
    "newimages2=[]\n",
    "for i in range(4):\n",
    "    new1 = new_images(image=recons_embed[0],channel=i)\n",
    "    newimages.append(new1)\n",
    "    new1 = new_images(image=recons_embed[1],channel=i)\n",
    "    newimages2.append(new1)\n",
    "    \n",
    "#can insert scaling code here\n",
    "\n",
    "fig, arrax = plt.subplots(2, 4, figsize=(12,6))\n",
    "for i in range(4):\n",
    "    arrax[0,i].imshow(newimages[i],cmap=cmap)\n",
    "    arrax[1,i].imshow(newimages2[i],cmap=cmap)\n",
    "    arrax[0,i].axis(\"off\")\n",
    "    arrax[1,i].axis(\"off\")\n",
    "arrax[0,0].set_title(\"Generated Images (channels 1-4)\")\n",
    "arrax[1,0].set_title(\"Generated Images (channels 1-4)\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f2a718f7-d9a1-4df2-bda2-b8c810428f2e",
   "metadata": {},
   "source": [
    "## Save Convolutional Autoencoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "7b7d6ca5-7484-4f78-a6c0-0f3ebd834859",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: 6731_30March_test\\assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: 6731_30March_test\\assets\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Autoencoder model saved.\n",
      "WARNING:tensorflow:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: 6731_30March_test_encoder\\assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: 6731_30March_test_encoder\\assets\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Encoder model saved.\n",
      "WARNING:tensorflow:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: 6731_30March_test_decoder\\assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: 6731_30March_test_decoder\\assets\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decoder model saved.\n"
     ]
    }
   ],
   "source": [
    "autoencoder.save(\"6731_31March_test\",save_format=\"tf\")\n",
    "print(\"Autoencoder model saved.\")\n",
    "\n",
    "encoder.save(\"6731_31March_test_encoder\",save_format=\"tf\")\n",
    "print(\"Encoder model saved.\")\n",
    "\n",
    "decoder.save(\"6731_31March_test_decoder\",save_format=\"tf\")\n",
    "print(\"Decoder model saved.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96afa96f-6948-4c9a-add7-225ff943c148",
   "metadata": {},
   "source": [
    "## Load and restart training on the Convolutional Autoencoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "8c5f08b6-39cd-4671-a73f-fb30ac584c73",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 61/61\n",
      "INFO:tensorflow:Assets written to: ConvAE\\ConvAE\\assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: ConvAE\\ConvAE\\assets\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "190/190 - 43s - loss: 0.0554 - mse: 0.0517 - val_loss: 0.0538 - val_mse: 0.0510 - 43s/epoch - 226ms/step\n"
     ]
    }
   ],
   "source": [
    "autoencoder = tf.keras.models.load_model(\"6731_30March_test\")\n",
    "epcs = 60\n",
    "\n",
    "optimizer = tf.keras.optimizers.Adam(learning_rate=0.001)\n",
    "autoencoder.compile(optimizer=optimizer,metrics=[\"mse\"])\n",
    "checkpoint_cb = tf.keras.callbacks.ModelCheckpoint(r\"ConvAE/ConvAE\",monitor=\"val_mse\",save_best_only=True)\n",
    "\n",
    "history = autoencoder.fit(dataset.X,dataset.X,epochs=epcs+1,verbose=2,validation_split=0.1,initial_epoch=epcs,callbacks=[checkpoint_cb])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "344f591e-76aa-4f78-803e-9acd15ab7ac7",
   "metadata": {},
   "source": [
    "## Visualize embedded distributions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4d7ddc14-30f2-4a98-9990-61e7bdb9a10a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 2s 2s/step\n",
      "Plot a sample distribution from the input: \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 8 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "\n",
    "#load a model\n",
    "#encoder = tf.keras.models.load_model(\"6731_28March_50EPCS_encoder\")\n",
    "\n",
    "cmap=\"tab20c\"\n",
    "\n",
    "u1,u2,z_plot = encoder.predict(dataset.X[:8])\n",
    "print(\"Plot a sample distribution from the input: \")\n",
    "\n",
    "fig, arrax = plt.subplots(2, 4, figsize=(12,6))\n",
    "for i in range(2):\n",
    "    for j in range(4):\n",
    "        arrax[i,j].hist(z_plot[i*4+j])\n",
    "        arrax[i,j].axis(\"off\")\n",
    "        \n",
    "arrax[0,0].set_title(\"Sample distributions from embedded layer\")\n",
    "arrax[1,0].set_title(\"Sample distributions from embedded layer\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5e80d75-0a3a-4ddf-adc3-f8f1e94088c6",
   "metadata": {},
   "source": [
    "## Code for scaling images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "68654c36-66de-49fe-a5dc-b54b51b01c13",
   "metadata": {},
   "outputs": [],
   "source": [
    "#scaling for images\n",
    "biggest, smallest = 0.0, 6000.0\n",
    "tmp=np.empty(2)\n",
    "for i in range(4):\n",
    "    tmp[0] = newimages[i].max()\n",
    "    tmp[1] = oldimages[i].max()\n",
    "    tmps = tmp.max()\n",
    "    if tmps > biggest:\n",
    "        biggest = tmps\n",
    "    tmp[0] = newimages[i].min()\n",
    "    tmp[1] = oldimages[i].min()\n",
    "    tmps = tmp.min()\n",
    "    if tmps < smallest:\n",
    "        smallest = tmps\n",
    "scaled_newimages = ((newimages - smallest) * (1/(biggest-smallest)) * 255).astype('uint8')\n",
    "scaled_oldimages = ((oldimages - smallest) * (1/(biggest-smallest)) * 255).astype('uint8')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}