jp-workbook.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## this is JP's experimental workbook.\n",
    "## Please do not modify this file - you are welcome to save a copy\n",
    "## for yourself"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import os, sys, glob\n",
    "import numpy as np\n",
    "import scipy as sp\n",
    "import copy\n",
    "import keras\n",
    "from graphClasses import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## layers import\n",
    "from keras.layers import SimpleRNN, LSTM, Input, Dense, RNN, SimpleRNNCell\n",
    "from keras.models import Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "105"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import itertools as it\n",
    "cc = it.combinations(range(0,15),2)\n",
    "len(list(cc))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "graphlist = glob.glob('qm9graph/*.csv')\n",
    "graphs = []\n",
    "for g in graphlist:\n",
    "    g1 = simpleGraph()\n",
    "    g1.graphLoader(g)\n",
    "    graphs.append(g1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 1., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.],\n",
       "       [1., 0., 1., 0., 0., 0., 1., 0., 0., 0., 1., 0., 0., 0., 0.],\n",
       "       [0., 1., 0., 1., 0., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0.],\n",
       "       [0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 1., 0., 0.],\n",
       "       [0., 0., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 1., 0.],\n",
       "       [0., 1., 1., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 0., 0., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 1.],\n",
       "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.]])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "g1.am"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1, 9]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "g1.getNeighbors1(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0, 2, 6, 10]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "g1.getNeighbors1(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[2, 6, 10]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "g1.getNeighbors2(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "210\n",
      "225\n"
     ]
    }
   ],
   "source": [
    "ind = 14\n",
    "maxdim = 15\n",
    "\n",
    "print(int(ind*maxdim))\n",
    "print(int(ind*maxdim + maxdim))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "xx = np.zeros((len(graphs),105))\n",
    "yy = np.zeros((len(graphs),15))\n",
    "for ii, gr in enumerate(graphs):\n",
    "    xx[ii,:] = np.reshape(gr.edges,(1,105))\n",
    "    yy[ii,:] = np.reshape(gr.ats,(1,15))\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 1., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.],\n",
       "       [1., 0., 1., 0., 0., 0., 1., 0., 0., 0., 1., 0., 0., 0., 0.],\n",
       "       [0., 1., 0., 1., 0., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0.],\n",
       "       [0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 1., 0., 0.],\n",
       "       [0., 0., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 1., 0.],\n",
       "       [0., 1., 1., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 0., 0., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 1.],\n",
       "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.]])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "g1.am"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5, 5)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.zeros((5,5)).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'qm9graph/021349.csv'"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "graphlist[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## let's a sequence classifier\n",
    "maxdim = 15\n",
    "inputs = Input(shape=(15,1))\n",
    "LSTMstack1 = LSTM(units=100,return_sequences=True)(inputs)\n",
    "LSTMstack2 = LSTM(units=100,return_sequences=True)(LSTMstack1)\n",
    "output = Dense(units=6,activation='softmax')(LSTMstack2)\n",
    "LSTM_classifier_model = Model(inputs=inputs, outputs=output)\n",
    "LSTM_classifier_model.compile(optimizer='adam',loss='categorical_crossentropy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## let's try a simply fully connected network\n",
    "maxdim = 15\n",
    "inputs = Input(shape=(105,))\n",
    "\n",
    "dense1 = Dense(units=100,activation='relu')(inputs)\n",
    "dense2 = Dense(units=100,activation='relu')(dense1)\n",
    "output = Dense(units=15,activation='linear')(dense2)\n",
    "\n",
    "model = Model(inputs=inputs, outputs=output)\n",
    "model.compile(optimizer='adam',loss='mean_squared_error')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "193/193 [==============================] - 1s 6ms/step - loss: 1.1465\n",
      "Epoch 2/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 1.0985\n",
      "Epoch 3/20\n",
      "193/193 [==============================] - 1s 6ms/step - loss: 1.0510\n",
      "Epoch 4/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 1.0034\n",
      "Epoch 5/20\n",
      "193/193 [==============================] - 2s 8ms/step - loss: 0.9735\n",
      "Epoch 6/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 0.9205\n",
      "Epoch 7/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 0.8819\n",
      "Epoch 8/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 0.8465\n",
      "Epoch 9/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 0.8189\n",
      "Epoch 10/20\n",
      "193/193 [==============================] - 1s 6ms/step - loss: 0.7693\n",
      "Epoch 11/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 0.7426\n",
      "Epoch 12/20\n",
      "193/193 [==============================] - 1s 6ms/step - loss: 0.7060\n",
      "Epoch 13/20\n",
      "193/193 [==============================] - 1s 6ms/step - loss: 0.6785\n",
      "Epoch 14/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 0.6576\n",
      "Epoch 15/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 0.6431\n",
      "Epoch 16/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 0.6201\n",
      "Epoch 17/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 0.6087\n",
      "Epoch 18/20\n",
      "193/193 [==============================] - 2s 8ms/step - loss: 0.5884\n",
      "Epoch 19/20\n",
      "193/193 [==============================] - 1s 7ms/step - loss: 0.5731\n",
      "Epoch 20/20\n",
      "193/193 [==============================] - 1s 6ms/step - loss: 0.5630\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7f568c0c8750>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(xx,yy,epochs=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(193, 15, 6)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yyy.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## sequence X values\n",
    "xxx = np.reshape(xx,(len(graphs),105,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## form the y values \n",
    "from sklearn.preprocessing import OneHotEncoder,LabelEncoder\n",
    "ohe_encoder = OneHotEncoder()\n",
    "yyy= np.zeros([len(graphs),15,6])\n",
    "yyyy = np.reshape(yy,(len(graphs),15,1))\n",
    "for i, y in  enumerate(yy):\n",
    "    for j,e in enumerate(y):\n",
    "        if int(e) == 0:\n",
    "            yyy[i,j,0] = 1\n",
    "        elif int(e) == 1:\n",
    "            yyy[i,j,1] = 1\n",
    "        elif int(e) == 6:\n",
    "            yyy[i,j,2] = 1\n",
    "        elif int(e) == 7:\n",
    "            yyy[i,j,3] = 1\n",
    "        elif int(e) == 8:\n",
    "            yyy[i,j,4] = 1\n",
    "        elif int(e) == 9:\n",
    "            yyy[i,j,5] = 1\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model.fit(x=xx,y=yyy,epochs=5,batch_size=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "p = LSTM_classifier_model.predict(yyyy)\n",
    "p.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'LSTM_classifier_model' 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-27-520661bd7774>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mLSTM_classifier_model\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0myyyy\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0my\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0myyy\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mepochs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mbatch_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m: name 'LSTM_classifier_model' is not defined"
     ]
    }
   ],
   "source": [
    "LSTM_classifier_model.fit(x=yyyy,y=yyy,epochs=5,batch_size=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "p = LSTM_classifier_model.predict(yyyy)\n",
    "p.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "p[2][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "yyy[2][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from IPython.display import SVG\n",
    "from keras.utils.vis_utils import model_to_dot\n",
    "\n",
    "SVG(model_to_dot(model,show_shapes=True).create(prog='dot', format='svg'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "gh.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "w1 = model.get_weights()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "w1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "[i.shape for i in w1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model.set_weights(weights=[np.reshape(np.array([1.0]),(1,1)),\n",
    "                           np.reshape(np.array([1.0]),(1,1)),\n",
    "                           np.array([1.0])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "w1 = model.get_weights()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "w1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "g1.edges.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "gh.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "hide_input": false,
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}