keras_linear_regression.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Linear regression in sklearn v.s. Keras. \n",
    "- Validation data in Keras\n",
    "- Getting coefficients (weights) in Keras."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using Theano backend.\n"
     ]
    }
   ],
   "source": [
    "import ds_utils.imports; import imp; imp.reload(ds_utils.imports)\n",
    "from ds_utils.imports import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X = np.random.uniform(size=1000).reshape(-1, 1)\n",
    "bias = 1\n",
    "w = 2\n",
    "noise = np.random.normal(scale=0.1, size=y.shape)\n",
    "y = np.dot(X, w) + bias + noise\n",
    "y = y.reshape(-1, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7fcd23196f60>"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrubrCLj6UqYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fcd2321c320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = sklearn.model_selection.train_test_split(\n",
    "    X, y, test_size=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((800, 1), (800, 1), (200, 1), (200, 1))"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, y_train.shape, X_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Linear regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "mdl = sklearn.linear_model.LinearRegression().fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y_pred = mdl.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.010623054344415111"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sklearn.metrics.mean_squared_error(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.98434475]])"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mdl.coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7fcd230cdd68>"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npOAuIlKFFNxFRKrQ/wFJt90b038t8AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fcd230cdc88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X_test, y_test)\n",
    "plt.scatter(X_test, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Keras linear regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 800 samples, validate on 200 samples\n",
      "Epoch 1/10\n",
      "800/800 [==============================] - 0s - loss: 0.2397 - val_loss: 0.0170\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\n",
      "Epoch 2/10\n",
      "800/800 [==============================] - 0s - loss: 0.0141 - val_loss: 0.0141\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\n",
      "Epoch 3/10\n",
      "800/800 [==============================] - 0s - loss: 0.0121 - val_loss: 0.0125\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\n",
      "Epoch 4/10\n",
      "800/800 [==============================] - 0s - loss: 0.0111 - val_loss: 0.0118\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\n",
      "Epoch 5/10\n",
      "800/800 [==============================] - 0s - loss: 0.0105 - val_loss: 0.0112\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\n",
      "Epoch 6/10\n",
      "800/800 [==============================] - 0s - loss: 0.0102 - val_loss: 0.0112\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\n",
      "Epoch 7/10\n",
      "800/800 [==============================] - 0s - loss: 0.0101 - val_loss: 0.0108\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\n",
      "Epoch 8/10\n",
      "800/800 [==============================] - 0s - loss: 0.0101 - val_loss: 0.0108\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\n",
      "Epoch 9/10\n",
      "800/800 [==============================] - 0s - loss: 0.0100 - val_loss: 0.0107\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\n",
      "Epoch 10/10\n",
      "800/800 [==============================] - 0s - loss: 0.0099 - val_loss: 0.0112\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7fcd22dd76d8>"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mdl = keras.models.Sequential(\n",
    "    [keras.layers.Dense(\n",
    "        units=y_train.shape[1], input_dim=X_train.shape[1])])\n",
    "mdl.compile(optimizer=keras.optimizers.SGD(lr=0.1), loss='mse', metrics=[])\n",
    "mdl.fit(x=X_train,\n",
    "        y=y_train,\n",
    "        epochs=10,\n",
    "        batch_size=32,\n",
    "        validation_data=(X_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "y_pred = mdl.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.01116122625907987"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sklearn.metrics.mean_squared_error(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7fcd22d205c0>"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fcd22cffa20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X_test, y_test)\n",
    "plt.scatter(X_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      " 32/200 [===>..........................] - ETA: 0s"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.011161224450916052"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mdl.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_6 (Dense)              (None, 1)                 2         \n",
      "=================================================================\n",
      "Total params: 2.0\n",
      "Trainable params: 2\n",
      "Non-trainable params: 0.0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "mdl.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([[ 1.96372533]], dtype=float32), array([ 1.00714719], dtype=float32)]"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mdl.get_layer('dense_6').get_weights()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([[ 1.96372533]], dtype=float32), array([ 1.00714719], dtype=float32)]"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mdl.get_weights()"
   ]
  }
 ],
 "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": "3.6.0"
  },
  "toc": {
   "colors": {
    "hover_highlight": "#DAA520",
    "running_highlight": "#FF0000",
    "selected_highlight": "#FFD700"
   },
   "moveMenuLeft": true,
   "nav_menu": {
    "height": "65px",
    "width": "251px"
   },
   "navigate_menu": true,
   "number_sections": true,
   "sideBar": true,
   "threshold": 4,
   "toc_cell": false,
   "toc_section_display": "block",
   "toc_window_display": true,
   "widenNotebook": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}