Keras example.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Keras example for CPOL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using Theano backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "## You are using the Python ARM Radar Toolkit (Py-ART), an open source\n",
      "## library for working with weather radar data. Py-ART is partly\n",
      "## supported by the U.S. Department of Energy as part of the Atmospheric\n",
      "## Radiation Measurement (ARM) Climate Research Facility, an Office of\n",
      "## Science user facility.\n",
      "##\n",
      "## If you use this software to prepare a publication, please cite:\n",
      "##\n",
      "##     JJ Helmus and SM Collis, JORS 2016, doi: 10.5334/jors.119\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/rjackson/anaconda3/envs/pyart_ml/lib/python3.7/site-packages/pyart/graph/cm.py:104: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
      "  if 'red' in spec:\n",
      "/home/rjackson/anaconda3/envs/pyart_ml/lib/python3.7/site-packages/pyart/graph/cm_colorblind.py:32: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
      "  if 'red' in spec:\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential, load_model\n",
    "from keras.layers import Dense, Activation, Embedding, Flatten, Dropout, TimeDistributed, Reshape, Lambda\n",
    "from keras.layers import LSTM, ConvLSTM2D, Conv2D, ZeroPadding2D\n",
    "from keras.optimizers import RMSprop, Adam, SGD\n",
    "from keras import backend as K\n",
    "from keras.utils import to_categorical, Sequence\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.layers.normalization import BatchNormalization\n",
    "import keras\n",
    "import pyart\n",
    "import numpy as np\n",
    "from glob import glob\n",
    "import matplotlib.pyplot as plt\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_greyscale_field(radar_file, num_ref_levels=64):\n",
    "    grid = pyart.io.read_grid(radar_file, include_fields='reflectivity')\n",
    "    refl = grid.fields['reflectivity']['data']\n",
    "    refl = np.ma.filled(refl, np.nan)\n",
    "    #refl[np.isnan(refl)] = 0\n",
    "    comp_refl = np.nanmax(refl, axis=0)\n",
    "    comp_refl[np.isnan(comp_refl)] = 0\n",
    "    # We will not care about < 0 dBZ since we are targeting precipitation events\n",
    "    #comp_refl[comp_refl < 0] = 0\n",
    "    del grid\n",
    "    if comp_refl.max() > 0:\n",
    "        my_vals = comp_refl/comp_refl.max()\n",
    "    else:\n",
    "        my_vals = np.zeros_like(comp_refl)\n",
    "    return my_vals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4460\n"
     ]
    }
   ],
   "source": [
    "radar_list = glob('/lcrc/group/earthscience/radar/CPOL_level_1b/GRIDDED/GRID_150km_2500m/2006/200601**/*', recursive=True)\n",
    "#radar_list = radar_list + glob('/home/rjackson/data/radar/gridded_data/20060121/*1000m*.nc', recursive=True)\n",
    "radar_list = sorted(radar_list)\n",
    "print(len(radar_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(117, 117)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "first_greyscale_field = get_greyscale_field(radar_list[0])\n",
    "first_greyscale_field.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x2b62b40ef1d0>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(first_greyscale_field, cmap='pyart_HomeyerRainbow')\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "class DataGenerator(Sequence):\n",
    "    'Generates data for Keras'\n",
    "    def __init__(self, list_IDs, labels, batch_size=32, dim=(117,117), n_channels=1,\n",
    "                 n_classes=10, shuffle=False):\n",
    "        'Initialization'\n",
    "        self.dim = dim\n",
    "        self.batch_size = batch_size\n",
    "        self.labels = labels\n",
    "        self.list_IDs = list_IDs\n",
    "        self.n_channels = n_channels\n",
    "        self.n_classes = n_classes\n",
    "        self.shuffle = shuffle\n",
    "        self.on_epoch_end()\n",
    "\n",
    "    def __len__(self):\n",
    "        'Denotes the number of batches per epoch'\n",
    "        return int(np.floor(len(self.list_IDs) / self.batch_size))\n",
    "\n",
    "    def __getitem__(self, index):\n",
    "        'Generate one batch of data'\n",
    "        # Generate indexes of the batch\n",
    "        indexes = self.indexes[index*self.batch_size:(index+1)*self.batch_size]\n",
    "\n",
    "        # Find list of IDs\n",
    "        list_IDs_temp = [self.list_IDs[k] for k in indexes]\n",
    "\n",
    "        # Generate data\n",
    "        X, y = self.__data_generation(list_IDs_temp)\n",
    "\n",
    "        return X, y\n",
    "\n",
    "    def on_epoch_end(self):\n",
    "        'Updates indexes after each epoch'\n",
    "        self.indexes = np.arange(len(self.list_IDs))\n",
    "        if self.shuffle == True:\n",
    "            np.random.shuffle(self.indexes)\n",
    "\n",
    "    def __data_generation(self, list_IDs_temp):\n",
    "        'Generates data containing batch_size samples' # X : (n_samples, *dim, n_channels)\n",
    "        # Initialization\n",
    "        X = np.empty((self.batch_size, 3, *self.dim, self.n_channels))\n",
    "        y = np.empty((self.batch_size, *self.dim, 1))\n",
    "\n",
    "        # Generate data\n",
    "        for i, ID in enumerate(list_IDs_temp):\n",
    "            # Store sample\n",
    "            for j in range(3):\n",
    "                X[i, j, :, :, 0] = get_greyscale_field(radar_list[ID+j])\n",
    "                # Store class\n",
    "            the_array = get_greyscale_field(radar_list[ID+3])\n",
    "            y[i, :, :, 0] = the_array\n",
    "        return X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1494.3008\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-100-3297f3e81ae8>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtraining_data\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m<ipython-input-98-8f907ea3614f>\u001b[0m in \u001b[0;36m__getitem__\u001b[0;34m(self, index)\u001b[0m\n\u001b[1;32m     26\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     27\u001b[0m         \u001b[0;31m# Generate data\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 28\u001b[0;31m         \u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__data_generation\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlist_IDs_temp\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     29\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     30\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-98-8f907ea3614f>\u001b[0m in \u001b[0;36m__data_generation\u001b[0;34m(self, list_IDs_temp)\u001b[0m\n\u001b[1;32m     48\u001b[0m                 \u001b[0mX\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_greyscale_field\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mradar_list\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mID\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0mj\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     49\u001b[0m                 \u001b[0;31m# Store class\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 50\u001b[0;31m             \u001b[0mthe_array\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_greyscale_field\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mradar_list\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mID\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;36m3\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     51\u001b[0m             \u001b[0my\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mthe_array\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     52\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-3-7e044017fe7e>\u001b[0m in \u001b[0;36mget_greyscale_field\u001b[0;34m(radar_file, num_ref_levels)\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mget_greyscale_field\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mradar_file\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnum_ref_levels\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m64\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m     \u001b[0mgrid\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpyart\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mio\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread_grid\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mradar_file\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minclude_fields\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'reflectivity'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      3\u001b[0m     \u001b[0mrefl\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgrid\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfields\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'reflectivity'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'data'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m     \u001b[0mrefl\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mma\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfilled\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrefl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnan\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m     \u001b[0;31m#refl[np.isnan(refl)] = 0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.local/lib/python3.6/site-packages/pyart/io/grid_io.py\u001b[0m in \u001b[0;36mread_grid\u001b[0;34m(filename, exclude_fields, **kwargs)\u001b[0m\n\u001b[1;32m     96\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mfield\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mexclude_fields\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     97\u001b[0m             \u001b[0;32mcontinue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 98\u001b[0;31m         \u001b[0mfield_dic\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_ncvar_to_dict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdset\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvariables\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mfield\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     99\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mfield_dic\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'data'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mfield_shape_with_time\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    100\u001b[0m             \u001b[0mfield_dic\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'data'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfield_shape\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.local/lib/python3.6/site-packages/pyart/io/cfradial.py\u001b[0m in \u001b[0;36m_ncvar_to_dict\u001b[0;34m(ncvar, lazydict)\u001b[0m\n\u001b[1;32m    353\u001b[0m         \u001b[0md\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_lazy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'data'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata_extractor\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    354\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 355\u001b[0;31m         \u001b[0md\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'data'\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdata_extractor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    356\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0md\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    357\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.local/lib/python3.6/site-packages/pyart/io/cfradial.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    374\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    375\u001b[0m         \u001b[0;34m\"\"\" Return an array containing data from the stored variable. \"\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 376\u001b[0;31m         \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mncvar\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    377\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mdata\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mma\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmasked\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    378\u001b[0m             \u001b[0;31m# If the data is a masked scalar, MaskedConstant is returned by\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32mnetCDF4/_netCDF4.pyx\u001b[0m in \u001b[0;36mnetCDF4._netCDF4.Variable.__getitem__\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mnetCDF4/_netCDF4.pyx\u001b[0m in \u001b[0;36mnetCDF4._netCDF4.Variable.shape.__get__\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32m/blues/gpfs/home/rjackson/anaconda3/envs/yt_devel/lib/python3.6/site-packages/netCDF4/utils.py\u001b[0m in \u001b[0;36m_find_dim\u001b[0;34m(grp, dimname)\u001b[0m\n\u001b[1;32m     38\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mA\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0msorted\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mA\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mB\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__getitem__\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     39\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 40\u001b[0;31m \u001b[0;32mdef\u001b[0m \u001b[0m_find_dim\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mgrp\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdimname\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     41\u001b[0m     \u001b[0;31m# find Dimension instance given group and name.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     42\u001b[0m     \u001b[0;31m# look in current group, and parents.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1533.3872\n",
      "1515.1788\n"
     ]
    }
   ],
   "source": [
    "x, y = training_data[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       ...,\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.]])"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y[8,:,:,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.QuadMesh at 0x2b7825a9c080>"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.pcolormesh(y[12,:,:,0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'x' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-8-39f05a93cc11>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpcolormesh\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m: name 'x' is not defined"
     ]
    }
   ],
   "source": [
    "plt.pcolormesh(x[0][0,1,:,:,0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_steps = 3\n",
    "conv_window = 3\n",
    "batch_size = 1\n",
    "train_data_len = round(0.8*len(radar_list))\n",
    "labels = {}\n",
    "# Create labels for each member of radar dataset\n",
    "for i in range(len(radar_list)):\n",
    "    label_key = \"id-\" + str(i)\n",
    "    labels[label_key] = i\n",
    "    \n",
    "training_data = DataGenerator([int(x) for x in range(0, train_data_len)], labels,\n",
    "                              batch_size=30, n_classes=1)\n",
    "validation_data = DataGenerator([int(x) for x in range(train_data_len, len(radar_list))], labels,\n",
    "                                batch_size=30, n_classes=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO (theano.gof.compilelock): Refreshing lock /blues/gpfs/home/rjackson/.theano/compiledir_Linux-3.10-el7.x86_64-x86_64-with-centos-7.6.1810-Core-x86_64-3.7.3-64/lock_dir/lock\n"
     ]
    }
   ],
   "source": [
    "hidden_size = 5\n",
    "use_dropout=True\n",
    "the_array = first_greyscale_field\n",
    "new_shape = (num_steps, the_array.shape[0], the_array.shape[1], 1)\n",
    "new_shape_conv10 = (num_steps, the_array.shape[0]-conv_window, the_array.shape[1]-conv_window, 1)\n",
    "model = Sequential()\n",
    "\n",
    "model.add(ConvLSTM2D(filters=1, kernel_size=(9, 9),\n",
    "                     input_shape=new_shape, data_format='channels_last', return_sequences=True,\n",
    "                     padding='valid', activation='relu'))\n",
    "model.add(ConvLSTM2D(filters=1, kernel_size=(conv_window, conv_window),\n",
    "                     input_shape=new_shape, data_format='channels_last', return_sequences=False,\n",
    "                     padding='valid', activation='relu'))\n",
    "\n",
    "model.add(ZeroPadding2D(5))\n",
    "#odel.add(ConvLSTM2D(1, 10, input_shape=new_shape_conv10, data_format='channels_last', return_sequences=True))\n",
    "#if use_dropout:\n",
    "#    model.add(Dropout(0.5))\n",
    "model.compile(loss='mean_squared_error', optimizer='adam')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_path = '/home/rjackson/deep_learning_fun'\n",
    "checkpointer = ModelCheckpoint(filepath=data_path + '/2-layer-model-{epoch:02d}.hdf5', verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "118/118 [==============================] - 4516s 38s/step - loss: 0.0144\n",
      "\n",
      "Epoch 00001: saving model to /home/rjackson/deep_learning_fun/2-layer-model-01.hdf5\n",
      "Epoch 2/10\n",
      "118/118 [==============================] - 4543s 39s/step - loss: 0.0134\n",
      "\n",
      "Epoch 00002: saving model to /home/rjackson/deep_learning_fun/2-layer-model-02.hdf5\n",
      "Epoch 3/10\n",
      "118/118 [==============================] - 4568s 39s/step - loss: 0.0128\n",
      "\n",
      "Epoch 00003: saving model to /home/rjackson/deep_learning_fun/2-layer-model-03.hdf5\n",
      "Epoch 4/10\n",
      "118/118 [==============================] - 7559s 64s/step - loss: 0.0125\n",
      "\n",
      "Epoch 00004: saving model to /home/rjackson/deep_learning_fun/2-layer-model-04.hdf5\n",
      "Epoch 5/10\n",
      " 14/118 [==>...........................] - ETA: 1:13:56 - loss: 0.0124"
     ]
    }
   ],
   "source": [
    "num_epochs = 30\n",
    "model.fit_generator(training_data, None, epochs=10, callbacks=[checkpointer])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO (theano.gof.compilelock): Waiting for existing lock by process '100463' (I am process '240758')\n",
      "INFO (theano.gof.compilelock): To manually release the lock, delete /blues/gpfs/home/rjackson/.theano/compiledir_Linux-3.10-el7.x86_64-x86_64-with-centos-7.6.1810-Core-x86_64-3.7.3-64/lock_dir\n"
     ]
    }
   ],
   "source": [
    "from keras.models import load_model\n",
    "model = load_model('/home/rjackson/deep_learning_fun/2-layer-model-03.hdf5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_model = model.predict_generator(validation_data, steps=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(900, 117, 117, 1)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_model.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "output_png_path_model = '/home/rjackson/various_notebooks/deep_learning_quicklooks/model'\n",
    "for i in range(0, y_model.shape[0]):\n",
    "    fig, ax = plt.subplots(1,1, figsize=(10,10))\n",
    "    im = ax.imshow(y_model[i,:,:,0], vmin=0, vmax=1, cmap='pyart_HomeyerRainbow')\n",
    "    ax.set_title('Time step ' + str(i))\n",
    "    ax.set_xlabel('X [km]')\n",
    "    ax.set_ylabel('Y [km]')\n",
    "    plt.colorbar(im, label='Normalized reflectivity')\n",
    "    fig.savefig(output_png_path_model + '/' + str(i) + '.png')\n",
    "    plt.close(fig)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "output_png_path_observations = '/home/rjackson/various_notebooks/deep_learning_quicklooks/observations'\n",
    "for j in range(30):\n",
    "    y_validation = training_data[j]\n",
    "    my_data = y_validation[1]\n",
    "    for i in range(0, my_data.shape[0]):\n",
    "        fig, ax = plt.subplots(1,1, figsize=(10,10))\n",
    "        im = ax.imshow(my_data[i,:,:,0], vmin=0, vmax=1, cmap='pyart_HomeyerRainbow')\n",
    "        ax.set_title('Time step ' + str(j*30 + i))\n",
    "        plt.colorbar(im, label='Normalized reflectivity')\n",
    "        fig.savefig(output_png_path_observations + '/' + str(j*30 + i) + '.png')\n",
    "        plt.close(fig)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "ename": "IndexError",
     "evalue": "index 3 is out of bounds for axis 0 with size 2",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-30-e07e41cca9cf>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mimin\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mimax\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m     \u001b[0mthe_plot\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0max\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0mimin\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m//\u001b[0m\u001b[0;36m6\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0mimin\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m%\u001b[0m\u001b[0;36m6\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mimshow\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_model\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvmin\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvmax\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcmap\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'pyart_HomeyerRainbow'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 6\u001b[0;31m     \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolorbar\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mthe_plot\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0max\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0max\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m//\u001b[0m\u001b[0;36m6\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m%\u001b[0m\u001b[0;36m6\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlabel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'Reflectivity [% of maximum]'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      7\u001b[0m     \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtitle\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'Predicted reflectivity (normalized)'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mIndexError\u001b[0m: index 3 is out of bounds for axis 0 with size 2"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x720 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2,6, figsize=(30,10))\n",
    "imin = 20\n",
    "imax = 30\n",
    "for i in range(imin, imax):\n",
    "    the_plot = ax[(i-imin)//6,(i-imin)%6].imshow(y_model[i,:,:,0], vmin=0, vmax=1, cmap='pyart_HomeyerRainbow')\n",
    "    plt.colorbar(the_plot, ax=ax[i//6,i%6], label='Reflectivity [% of maximum]')\n",
    "    plt.title('Predicted reflectivity (normalized)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(2,6, figsize=(30,10))\n",
    "my_data = y_validation[1]\n",
    "imin = 20\n",
    "imax = 30\n",
    "for i in range(imin, imax):\n",
    "    my_img = np.squeeze(my_data[i])\n",
    "    the_plot = ax[(i-imin)//6,(i-imin)%6].imshow(my_img/my_img.max(), vmin=0, vmax=1, cmap='pyart_HomeyerRainbow')\n",
    "    plt.colorbar(the_plot, ax=ax[(i-imin)//6,(i-imin)%6], label='Reflectivity [% of maximum]')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.pcolormesh(y_model[3, :,:,0], vmin=0, vmax=0.2, cmap='pyart_HomeyerRainbow')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_validation = validation_data[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.pcolormesh(np.squeeze(y[1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 117, 117, 1)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_validation[1][1].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0, 0, 0, ..., 0, 0, 0],\n",
       "       [0, 0, 0, ..., 0, 0, 0],\n",
       "       [0, 0, 0, ..., 0, 0, 0],\n",
       "       ...,\n",
       "       [0, 0, 0, ..., 0, 0, 0],\n",
       "       [0, 0, 0, ..., 0, 0, 0],\n",
       "       [0, 0, 0, ..., 0, 0, 0]])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_img"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda env:pyart_ml]",
   "language": "python",
   "name": "conda-env-pyart_ml-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}