matplotlib_matrix_array_print.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Using matplotlib to print in matrix array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pickle\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import os\n",
    "import sys  \n",
    "sys.path.insert(0, '/home/pytorch_ashish/deep_learning/data/alt_text_data_all/models/')\n",
    "from scipy.stats import norm\n",
    "import pandas as pd\n",
    "import cv2\n",
    "from models.AE import Autoencoder\n",
    "from utils.loaders import load_model, ImageLabelLoader\n",
    "from keras.preprocessing.image import ImageDataGenerator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "INPUT_DIM = (128,128,3)\n",
    "BATCH_SIZE = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 5)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image_dataframe=pd.read_csv('/home/pytorch_ashish/deep_learning/data/alt_text_data_all/image_location_venky.csv')\n",
    "image_dataframe=image_dataframe[0:1]\n",
    "image_dataframe.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 1 validated image filenames.\n"
     ]
    }
   ],
   "source": [
    "data_gen = ImageDataGenerator(rescale=1./255)\n",
    "\n",
    "data_flow_dataframe = data_gen.flow_from_dataframe(dataframe = image_dataframe\n",
    "                                         , directory = None\n",
    "                                         , x_col='Path'\n",
    "                                         , target_size = INPUT_DIM[:2]\n",
    "                                         , batch_size = BATCH_SIZE\n",
    "                                         , shuffle = False\n",
    "                                         , class_mode = 'input'\n",
    "                                            )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "example_batch_2 = next(data_flow_dataframe)\n",
    "example_images_2 = example_batch_2[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 128, 128, 3)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "example_images_2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/pytorch_ashish/deep_learning/data/alt_text_data_all/models/color/AE-100epochs/\n"
     ]
    }
   ],
   "source": [
    "#Create two folder logs and Testing inside this folder\n",
    "RUN_FOLDER='/home/pytorch_ashish/deep_learning/data/alt_text_data_all/models/color/AE-100epochs/'\n",
    "print(RUN_FOLDER)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "ae = load_model(Autoencoder, RUN_FOLDER)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 128, 128, 3)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reconst_images.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /opt/conda/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:422: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x2160 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "a=[]\n",
    "fig = plt.figure(figsize=(30, 30))\n",
    "fig.subplots_adjust(hspace=0.4, wspace=0.4)\n",
    "k=1\n",
    "for i in range(3):\n",
    "    z_points_2 = ae.encoder.predict(example_images_2)\n",
    "    reconst_images = ae.decoder.predict(z_points_2)\n",
    "    img = reconst_images[0].squeeze()\n",
    "    sub = fig.add_subplot(2, 6, k)\n",
    "    k+=1\n",
    "    sub.axis('off')        \n",
    "    sub.imshow(img)\n",
    "    img2 = example_images_2[0].squeeze()\n",
    "    sub2 = fig.add_subplot(2, 6, k)\n",
    "    k+=1\n",
    "    sub2.axis('off')        \n",
    "    sub2.imshow(img2)\n",
    "    a.extend(z_points_2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "a=np.array(a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.]], dtype=float32)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "z=cosine_similarity(a[1].reshape(1,200),a[0].reshape(1,200))\n",
    "z"
   ]
  },
  {
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}