Z10_I_save_model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.0.0-rc1\n",
      "2.2.4-tf\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "print(tf.__version__)\n",
    "\n",
    "print(keras.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:6: ParserWarning: Falling back to the 'python' engine because the 'c' engine does not support regex separators (separators > 1 char and different from '\\s+' are interpreted as regex); you can avoid this warning by specifying engine='python'.\n",
      "  \n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:7: ParserWarning: Falling back to the 'python' engine because the 'c' engine does not support regex separators (separators > 1 char and different from '\\s+' are interpreted as regex); you can avoid this warning by specifying engine='python'.\n",
      "  import sys\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>workclass</th>\n",
       "      <th>fnlwgt</th>\n",
       "      <th>education</th>\n",
       "      <th>education_num</th>\n",
       "      <th>marital_status</th>\n",
       "      <th>occupation</th>\n",
       "      <th>relationship</th>\n",
       "      <th>race</th>\n",
       "      <th>sex</th>\n",
       "      <th>capital_gain</th>\n",
       "      <th>capital_loss</th>\n",
       "      <th>hours_per_week</th>\n",
       "      <th>native_country</th>\n",
       "      <th>wage_class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>39</td>\n",
       "      <td>State-gov</td>\n",
       "      <td>77516</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13</td>\n",
       "      <td>Never-married</td>\n",
       "      <td>Adm-clerical</td>\n",
       "      <td>Not-in-family</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>2174</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>United-States</td>\n",
       "      <td>&lt;=50K</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>50</td>\n",
       "      <td>Self-emp-not-inc</td>\n",
       "      <td>83311</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Exec-managerial</td>\n",
       "      <td>Husband</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>United-States</td>\n",
       "      <td>&lt;=50K</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>38</td>\n",
       "      <td>Private</td>\n",
       "      <td>215646</td>\n",
       "      <td>HS-grad</td>\n",
       "      <td>9</td>\n",
       "      <td>Divorced</td>\n",
       "      <td>Handlers-cleaners</td>\n",
       "      <td>Not-in-family</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>United-States</td>\n",
       "      <td>&lt;=50K</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>53</td>\n",
       "      <td>Private</td>\n",
       "      <td>234721</td>\n",
       "      <td>11th</td>\n",
       "      <td>7</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Handlers-cleaners</td>\n",
       "      <td>Husband</td>\n",
       "      <td>Black</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>United-States</td>\n",
       "      <td>&lt;=50K</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>28</td>\n",
       "      <td>Private</td>\n",
       "      <td>338409</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Prof-specialty</td>\n",
       "      <td>Wife</td>\n",
       "      <td>Black</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>40</td>\n",
       "      <td>Cuba</td>\n",
       "      <td>&lt;=50K</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age         workclass  fnlwgt  education  education_num  \\\n",
       "0   39         State-gov   77516  Bachelors             13   \n",
       "1   50  Self-emp-not-inc   83311  Bachelors             13   \n",
       "2   38           Private  215646    HS-grad              9   \n",
       "3   53           Private  234721       11th              7   \n",
       "4   28           Private  338409  Bachelors             13   \n",
       "\n",
       "       marital_status         occupation   relationship   race     sex  \\\n",
       "0       Never-married       Adm-clerical  Not-in-family  White    Male   \n",
       "1  Married-civ-spouse    Exec-managerial        Husband  White    Male   \n",
       "2            Divorced  Handlers-cleaners  Not-in-family  White    Male   \n",
       "3  Married-civ-spouse  Handlers-cleaners        Husband  Black    Male   \n",
       "4  Married-civ-spouse     Prof-specialty           Wife  Black  Female   \n",
       "\n",
       "   capital_gain  capital_loss  hours_per_week native_country wage_class  \n",
       "0          2174             0              40  United-States      <=50K  \n",
       "1             0             0              13  United-States      <=50K  \n",
       "2             0             0              40  United-States      <=50K  \n",
       "3             0             0              40  United-States      <=50K  \n",
       "4             0             0              40           Cuba      <=50K  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Wczytaj dane treningowe i testowe\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "\n",
    "train_set = pd.read_csv('Dane/adult/adult.data', sep=\", \",header = None)\n",
    "test_set = pd.read_csv('Dane/adult/adult.test', sep=\", \",skiprows = 1, header = None) # Make sure to skip a row for the test set\n",
    "\n",
    "col_labels = ['age', 'workclass', 'fnlwgt', 'education', 'education_num', 'marital_status', 'occupation', \n",
    "              'relationship', 'race', 'sex', 'capital_gain', 'capital_loss', 'hours_per_week', 'native_country',\n",
    "             'wage_class']\n",
    "train_set.columns = col_labels\n",
    "test_set.columns = col_labels\n",
    "\n",
    "train = train_set.replace('?', np.nan).dropna()\n",
    "test = test_set.replace('?', np.nan).dropna()\n",
    "\n",
    "train_set.head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Zad.\n",
    "Zróbmy szybki preprocesing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = pd.concat([train,test])\n",
    "\n",
    "dataset['wage_class'] = dataset.wage_class.replace({'<=50K.': 0,'<=50K':0, '>50K.':1, '>50K':1})\n",
    "\n",
    "dataset.drop([\"fnlwgt\"],axis=1,inplace=True)\n",
    "\n",
    "dataset.drop([\"education\"],axis=1,inplace=True)\n",
    "\n",
    "x = dataset.groupby('native_country')[\"wage_class\"].mean()\n",
    "\n",
    "d = dict(pd.cut(x[x.index!=\" United-States\"],5,labels=range(5)))\n",
    "\n",
    "dataset['native_country'] = dataset['native_country'].replace(d)\n",
    "\n",
    "dataset = pd.get_dummies(dataset,drop_first=True)\n",
    "\n",
    "train = dataset.iloc[:train.shape[0]]\n",
    "test = dataset.iloc[train.shape[0]:]\n",
    "\n",
    "X_train = train.drop(\"wage_class\",axis=1)\n",
    "y_train = train.wage_class\n",
    "\n",
    "X_test = test.drop(\"wage_class\",axis=1)\n",
    "y_test = test.wage_class\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.callbacks import ModelCheckpoint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_5 (Dense)              (None, 100)               4200      \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 50)                5050      \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              (None, 10)                510       \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 1)                 11        \n",
      "=================================================================\n",
      "Total params: 9,771\n",
      "Trainable params: 9,771\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "\n",
    "from keras.callbacks import History\n",
    "\n",
    "history = History()\n",
    "model = Sequential()\n",
    "model.add(Dense(100,activation=\"sigmoid\",input_shape=(X_train.shape[1],)))\n",
    "model.add(Dense(50,activation=\"sigmoid\"))\n",
    "model.add(Dense(10,activation=\"sigmoid\"))\n",
    "model.add(Dense(1,activation=\"sigmoid\"))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\ProgramData\\Anaconda3\\lib\\site-packages\\tensorflow_core\\python\\ops\\nn_impl.py:183: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n"
     ]
    }
   ],
   "source": [
    "save_best_model = ModelCheckpoint(\"wagi_best.h5py\",save_best_only=True)\n",
    "model.compile(loss=\"binary_crossentropy\",optimizer=\"Adam\", metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 30162 samples, validate on 15060 samples\n",
      "Epoch 1/100\n",
      "30162/30162 [==============================] - 2s 52us/step - loss: 0.4761 - accuracy: 0.7761 - val_loss: 0.3871 - val_accuracy: 0.8108\n",
      "Epoch 2/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3862 - accuracy: 0.8083 - val_loss: 0.3812 - val_accuracy: 0.8055\n",
      "Epoch 3/100\n",
      "30162/30162 [==============================] - 2s 51us/step - loss: 0.3754 - accuracy: 0.8134 - val_loss: 0.3770 - val_accuracy: 0.8155\n",
      "Epoch 4/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3718 - accuracy: 0.8134 - val_loss: 0.3654 - val_accuracy: 0.8177\n",
      "Epoch 5/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3723 - accuracy: 0.8116 - val_loss: 0.3715 - val_accuracy: 0.8168\n",
      "Epoch 6/100\n",
      "30162/30162 [==============================] - 1s 49us/step - loss: 0.3708 - accuracy: 0.8133 - val_loss: 0.3650 - val_accuracy: 0.8159\n",
      "Epoch 7/100\n",
      "30162/30162 [==============================] - 1s 49us/step - loss: 0.3698 - accuracy: 0.8127 - val_loss: 0.3939 - val_accuracy: 0.8047\n",
      "Epoch 8/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3702 - accuracy: 0.8150 - val_loss: 0.3895 - val_accuracy: 0.8065\n",
      "Epoch 9/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3681 - accuracy: 0.8136 - val_loss: 0.3665 - val_accuracy: 0.8112\n",
      "Epoch 10/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3663 - accuracy: 0.8149 - val_loss: 0.3653 - val_accuracy: 0.8153\n",
      "Epoch 11/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3653 - accuracy: 0.8173 - val_loss: 0.3786 - val_accuracy: 0.8158\n",
      "Epoch 12/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3667 - accuracy: 0.8156 - val_loss: 0.3645 - val_accuracy: 0.8162\n",
      "Epoch 13/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3648 - accuracy: 0.8138 - val_loss: 0.3678 - val_accuracy: 0.8117\n",
      "Epoch 14/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3648 - accuracy: 0.8142 - val_loss: 0.3643 - val_accuracy: 0.8137\n",
      "Epoch 15/100\n",
      "30162/30162 [==============================] - 1s 49us/step - loss: 0.3639 - accuracy: 0.8154 - val_loss: 0.3642 - val_accuracy: 0.8183\n",
      "Epoch 16/100\n",
      "30162/30162 [==============================] - 1s 49us/step - loss: 0.3646 - accuracy: 0.8162 - val_loss: 0.3781 - val_accuracy: 0.8106\n",
      "Epoch 17/100\n",
      "30162/30162 [==============================] - 2s 50us/step - loss: 0.3621 - accuracy: 0.8157 - val_loss: 0.3672 - val_accuracy: 0.8165\n",
      "Epoch 18/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3630 - accuracy: 0.8170 - val_loss: 0.3754 - val_accuracy: 0.8054\n",
      "Epoch 19/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3615 - accuracy: 0.8184 - val_loss: 0.3630 - val_accuracy: 0.8195\n",
      "Epoch 20/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3626 - accuracy: 0.8180 - val_loss: 0.3613 - val_accuracy: 0.8187\n",
      "Epoch 21/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3611 - accuracy: 0.8182 - val_loss: 0.3613 - val_accuracy: 0.8177\n",
      "Epoch 22/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3609 - accuracy: 0.8169 - val_loss: 0.3643 - val_accuracy: 0.8146\n",
      "Epoch 23/100\n",
      "30162/30162 [==============================] - 1s 49us/step - loss: 0.3626 - accuracy: 0.8159 - val_loss: 0.3656 - val_accuracy: 0.8151\n",
      "Epoch 24/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3609 - accuracy: 0.8167 - val_loss: 0.3636 - val_accuracy: 0.8175\n",
      "Epoch 25/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3606 - accuracy: 0.8179 - val_loss: 0.3723 - val_accuracy: 0.8150\n",
      "Epoch 26/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3604 - accuracy: 0.8173 - val_loss: 0.3649 - val_accuracy: 0.8144\n",
      "Epoch 27/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3611 - accuracy: 0.8173 - val_loss: 0.3649 - val_accuracy: 0.8139\n",
      "Epoch 28/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3599 - accuracy: 0.8184 - val_loss: 0.3785 - val_accuracy: 0.8116\n",
      "Epoch 29/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3603 - accuracy: 0.8173 - val_loss: 0.3604 - val_accuracy: 0.8181\n",
      "Epoch 30/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3593 - accuracy: 0.8200 - val_loss: 0.3614 - val_accuracy: 0.8187\n",
      "Epoch 31/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3594 - accuracy: 0.8194 - val_loss: 0.3683 - val_accuracy: 0.8118\n",
      "Epoch 32/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3598 - accuracy: 0.8183 - val_loss: 0.3660 - val_accuracy: 0.8189\n",
      "Epoch 33/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3596 - accuracy: 0.8170 - val_loss: 0.3666 - val_accuracy: 0.7942\n",
      "Epoch 34/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3596 - accuracy: 0.8186 - val_loss: 0.3613 - val_accuracy: 0.8204\n",
      "Epoch 35/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3588 - accuracy: 0.8180 - val_loss: 0.3626 - val_accuracy: 0.8148\n",
      "Epoch 36/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3585 - accuracy: 0.8197 - val_loss: 0.3763 - val_accuracy: 0.8155\n",
      "Epoch 37/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3583 - accuracy: 0.8195 - val_loss: 0.3662 - val_accuracy: 0.8109\n",
      "Epoch 38/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3595 - accuracy: 0.8179 - val_loss: 0.3632 - val_accuracy: 0.8195\n",
      "Epoch 39/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3579 - accuracy: 0.8217 - val_loss: 0.3627 - val_accuracy: 0.8167\n",
      "Epoch 40/100\n",
      "30162/30162 [==============================] - 1s 49us/step - loss: 0.3607 - accuracy: 0.8175 - val_loss: 0.3711 - val_accuracy: 0.8167\n",
      "Epoch 41/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3609 - accuracy: 0.8168 - val_loss: 0.3614 - val_accuracy: 0.8167\n",
      "Epoch 42/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3596 - accuracy: 0.8154 - val_loss: 0.3658 - val_accuracy: 0.8179\n",
      "Epoch 43/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3601 - accuracy: 0.8169 - val_loss: 0.3746 - val_accuracy: 0.8062\n",
      "Epoch 44/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3599 - accuracy: 0.8185 - val_loss: 0.3613 - val_accuracy: 0.8183\n",
      "Epoch 45/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3584 - accuracy: 0.8190 - val_loss: 0.3654 - val_accuracy: 0.8171\n",
      "Epoch 46/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3586 - accuracy: 0.8190 - val_loss: 0.3608 - val_accuracy: 0.8182\n",
      "Epoch 47/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3591 - accuracy: 0.8186 - val_loss: 0.3733 - val_accuracy: 0.8123\n",
      "Epoch 48/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3584 - accuracy: 0.8173 - val_loss: 0.3665 - val_accuracy: 0.8094\n",
      "Epoch 49/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3578 - accuracy: 0.8176 - val_loss: 0.3619 - val_accuracy: 0.8166\n",
      "Epoch 50/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3582 - accuracy: 0.8198 - val_loss: 0.3673 - val_accuracy: 0.8100\n",
      "Epoch 51/100\n",
      "30162/30162 [==============================] - 2s 50us/step - loss: 0.3578 - accuracy: 0.8193 - val_loss: 0.3626 - val_accuracy: 0.8191\n",
      "Epoch 52/100\n",
      "30162/30162 [==============================] - 1s 49us/step - loss: 0.3570 - accuracy: 0.8185 - val_loss: 0.3615 - val_accuracy: 0.8183\n",
      "Epoch 53/100\n",
      "30162/30162 [==============================] - 2s 55us/step - loss: 0.3572 - accuracy: 0.8210 - val_loss: 0.3635 - val_accuracy: 0.8141\n",
      "Epoch 54/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3576 - accuracy: 0.8182 - val_loss: 0.3630 - val_accuracy: 0.8165\n",
      "Epoch 55/100\n",
      "30162/30162 [==============================] - 1s 49us/step - loss: 0.3572 - accuracy: 0.8185 - val_loss: 0.3620 - val_accuracy: 0.8151\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 56/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3570 - accuracy: 0.8181 - val_loss: 0.3664 - val_accuracy: 0.8164\n",
      "Epoch 57/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3567 - accuracy: 0.8202 - val_loss: 0.3724 - val_accuracy: 0.8053\n",
      "Epoch 58/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3561 - accuracy: 0.8190 - val_loss: 0.3648 - val_accuracy: 0.8172\n",
      "Epoch 59/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3553 - accuracy: 0.8192 - val_loss: 0.3617 - val_accuracy: 0.8173\n",
      "Epoch 60/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3555 - accuracy: 0.8223 - val_loss: 0.3626 - val_accuracy: 0.8185\n",
      "Epoch 61/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3547 - accuracy: 0.8198 - val_loss: 0.3691 - val_accuracy: 0.8124\n",
      "Epoch 62/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3561 - accuracy: 0.8200 - val_loss: 0.3620 - val_accuracy: 0.8168\n",
      "Epoch 63/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3550 - accuracy: 0.8212 - val_loss: 0.3625 - val_accuracy: 0.8158\n",
      "Epoch 64/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3558 - accuracy: 0.8192 - val_loss: 0.3625 - val_accuracy: 0.8180\n",
      "Epoch 65/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3550 - accuracy: 0.8198 - val_loss: 0.3681 - val_accuracy: 0.8102\n",
      "Epoch 66/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3555 - accuracy: 0.8188 - val_loss: 0.3639 - val_accuracy: 0.8173\n",
      "Epoch 67/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3548 - accuracy: 0.8201 - val_loss: 0.3624 - val_accuracy: 0.8164\n",
      "Epoch 68/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3544 - accuracy: 0.8187 - val_loss: 0.3616 - val_accuracy: 0.8175\n",
      "Epoch 69/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3545 - accuracy: 0.8201 - val_loss: 0.3636 - val_accuracy: 0.8165\n",
      "Epoch 70/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3542 - accuracy: 0.8203 - val_loss: 0.3628 - val_accuracy: 0.8167\n",
      "Epoch 71/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3538 - accuracy: 0.8195 - val_loss: 0.3626 - val_accuracy: 0.8167\n",
      "Epoch 72/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3534 - accuracy: 0.8211 - val_loss: 0.3619 - val_accuracy: 0.8177\n",
      "Epoch 73/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3540 - accuracy: 0.8200 - val_loss: 0.3615 - val_accuracy: 0.8173\n",
      "Epoch 74/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3529 - accuracy: 0.8203 - val_loss: 0.3624 - val_accuracy: 0.8154\n",
      "Epoch 75/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3534 - accuracy: 0.8197 - val_loss: 0.3643 - val_accuracy: 0.8165\n",
      "Epoch 76/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3538 - accuracy: 0.8209 - val_loss: 0.3626 - val_accuracy: 0.8169\n",
      "Epoch 77/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3524 - accuracy: 0.8193 - val_loss: 0.3644 - val_accuracy: 0.8173\n",
      "Epoch 78/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3521 - accuracy: 0.8200 - val_loss: 0.3888 - val_accuracy: 0.7963\n",
      "Epoch 79/100\n",
      "30162/30162 [==============================] - 1s 49us/step - loss: 0.3529 - accuracy: 0.8196 - val_loss: 0.3623 - val_accuracy: 0.8169\n",
      "Epoch 80/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3520 - accuracy: 0.8217 - val_loss: 0.3647 - val_accuracy: 0.8141\n",
      "Epoch 81/100\n",
      "30162/30162 [==============================] - 2s 50us/step - loss: 0.3516 - accuracy: 0.8204 - val_loss: 0.3650 - val_accuracy: 0.8157\n",
      "Epoch 82/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3515 - accuracy: 0.8226 - val_loss: 0.3630 - val_accuracy: 0.8169\n",
      "Epoch 83/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3511 - accuracy: 0.8204 - val_loss: 0.3666 - val_accuracy: 0.8176\n",
      "Epoch 84/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3517 - accuracy: 0.8218 - val_loss: 0.3713 - val_accuracy: 0.8162\n",
      "Epoch 85/100\n",
      "30162/30162 [==============================] - 1s 49us/step - loss: 0.3513 - accuracy: 0.8222 - val_loss: 0.3638 - val_accuracy: 0.8165\n",
      "Epoch 86/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3510 - accuracy: 0.8208 - val_loss: 0.3628 - val_accuracy: 0.8163\n",
      "Epoch 87/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3511 - accuracy: 0.8212 - val_loss: 0.3682 - val_accuracy: 0.8162\n",
      "Epoch 88/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3506 - accuracy: 0.8216 - val_loss: 0.3673 - val_accuracy: 0.8129\n",
      "Epoch 89/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3507 - accuracy: 0.8220 - val_loss: 0.3631 - val_accuracy: 0.8161\n",
      "Epoch 90/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3517 - accuracy: 0.8218 - val_loss: 0.3652 - val_accuracy: 0.8165\n",
      "Epoch 91/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3509 - accuracy: 0.8227 - val_loss: 0.3642 - val_accuracy: 0.8155\n",
      "Epoch 92/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3497 - accuracy: 0.8202 - val_loss: 0.3658 - val_accuracy: 0.8160\n",
      "Epoch 93/100\n",
      "30162/30162 [==============================] - 1s 46us/step - loss: 0.3504 - accuracy: 0.8210 - val_loss: 0.3636 - val_accuracy: 0.8177\n",
      "Epoch 94/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3499 - accuracy: 0.8213 - val_loss: 0.3654 - val_accuracy: 0.8147\n",
      "Epoch 95/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3498 - accuracy: 0.8217 - val_loss: 0.3624 - val_accuracy: 0.8167\n",
      "Epoch 96/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3499 - accuracy: 0.8226 - val_loss: 0.3654 - val_accuracy: 0.8147\n",
      "Epoch 97/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3500 - accuracy: 0.8211 - val_loss: 0.3642 - val_accuracy: 0.8152\n",
      "Epoch 98/100\n",
      "30162/30162 [==============================] - 2s 51us/step - loss: 0.3497 - accuracy: 0.8217 - val_loss: 0.3668 - val_accuracy: 0.8152\n",
      "Epoch 99/100\n",
      "30162/30162 [==============================] - 1s 48us/step - loss: 0.3498 - accuracy: 0.8209 - val_loss: 0.3639 - val_accuracy: 0.8159\n",
      "Epoch 100/100\n",
      "30162/30162 [==============================] - 1s 47us/step - loss: 0.3496 - accuracy: 0.8216 - val_loss: 0.3630 - val_accuracy: 0.8159\n"
     ]
    }
   ],
   "source": [
    "save_best_model = ModelCheckpoint(\"wagi_best.h5py\",save_best_only=True)\n",
    "history = model.fit(X_train, y_train, validation_data= (X_test, y_test), batch_size=32,epochs=100, callbacks=[save_best_model])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "pd.DataFrame(history.history).plot(figsize=(8, 5))\n",
    "plt.grid(True)\n",
    "plt.gca().set_ylim(0, 1)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "15060/15060 [==============================] - 0s 16us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.36299147292437306, 0.8158698678016663]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8158698539176626"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import  metrics\n",
    "metrics.accuracy_score(y_true= y_test, y_pred= model.predict_classes(X_test) )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Zad\n",
    "\n",
    "Wczytaj nauczony model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "15060/15060 [==============================] - 0s 15us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.36044085541886006, 0.818061113357544]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.load_weights(\"wagi_best.h5py\")\n",
    "model.evaluate(X_test,y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Zad\n",
    "\n",
    "* Naucz dowolny model na zbiorze MNIST.\n",
    "* Zapisz optymalne parametry uczenia do pliku\n",
    "* Wczytaj dane z pliku"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}