word2vec/word2vec.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "### Importy"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:21:27.211216Z",
     "end_time": "2024-05-19T18:21:27.318205Z"
    }
   },
   "outputs": [],
   "source": [
    "import gzip\n",
    "import math\n",
    "import re\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from gensim.models import KeyedVectors\n",
    "from keras.layers import Dense, Dropout\n",
    "from keras.models import Sequential\n",
    "from keras.regularizers import l2\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Wczytywanie oraz czyszczenie danych"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "outputs": [],
   "source": [
    "def load_and_filter_data(file_path):\n",
    "    texts = []\n",
    "    labels = []\n",
    "    with gzip.open(file_path, 'rt', encoding='utf-8') as f:\n",
    "        for line in f:\n",
    "            parts = line.strip().split('\\t')\n",
    "            if len(parts) == 2:\n",
    "                labels.append(int(parts[0]))\n",
    "                texts.append(parts[1])\n",
    "    data = pd.DataFrame({'label': labels, 'text': texts})\n",
    "    return data\n",
    "\n",
    "def load_and_filter_tsv(file_path):\n",
    "    texts = []\n",
    "    with open(file_path, 'r', encoding='utf-8') as f:\n",
    "        for line in f:\n",
    "            parts = line.strip().split('\\t')\n",
    "            if len(parts) == 1:\n",
    "                texts.append(parts[0])\n",
    "    data = pd.DataFrame({'text': texts})\n",
    "    return data\n",
    "\n",
    "def load_labels(file_path):\n",
    "    labels = []\n",
    "    with open(file_path, 'r', encoding='utf-8') as f:\n",
    "        for line in f:\n",
    "            labels.append(int(line.strip()))\n",
    "    return np.array(labels)\n",
    "\n",
    "def clean_text(text):\n",
    "    text = text.lower()\n",
    "    text = re.sub(r'\\d+', '', text)\n",
    "    text = re.sub(r'\\s+', ' ', text)\n",
    "    text = re.sub(r'[^\\w\\s]', '', text)\n",
    "    return text"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:21:27.231204Z",
     "end_time": "2024-05-19T18:21:27.377342Z"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Wczytywanie danych treningowych oraz testowych"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "outputs": [],
   "source": [
    "train_data = load_and_filter_data('train/train.tsv.gz')\n",
    "train_data['text'] = train_data['text'].apply(clean_text)\n",
    "dev_data = load_and_filter_tsv('dev-0/in.tsv')\n",
    "dev_data['text'] = dev_data['text'].apply(clean_text)\n",
    "dev_labels = load_labels('dev-0/expected.tsv')\n",
    "test_data = load_and_filter_tsv('test-A/in.tsv')\n",
    "test_data['text'] = test_data['text'].apply(clean_text)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:21:27.241222Z",
     "end_time": "2024-05-19T18:21:31.160229Z"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Wczytywanie modelu word2vec"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "outputs": [],
   "source": [
    "word2vec_model = KeyedVectors.load(\"word2vec_100_3_polish.bin\")"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:21:31.161230Z",
     "end_time": "2024-05-19T18:21:54.895038Z"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Przekształcenie danych na wektory"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "outputs": [],
   "source": [
    "def text_to_vector(text, model):\n",
    "    words = text.split()\n",
    "    word_vecs = [model[word] for word in words if word in model]\n",
    "    return np.mean(word_vecs, axis=0) if len(word_vecs) > 0 else np.zeros(model.vector_size)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:21:54.900047Z",
     "end_time": "2024-05-19T18:21:54.909040Z"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "outputs": [],
   "source": [
    "X_train = np.array([text_to_vector(text, word2vec_model) for text in train_data['text']])\n",
    "y_train = np.array(train_data['label'])\n",
    "X_dev = np.array([text_to_vector(text, word2vec_model) for text in dev_data['text']])\n",
    "X_test = np.array([text_to_vector(text, word2vec_model) for text in test_data['text']])"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:21:54.913039Z",
     "end_time": "2024-05-19T18:22:03.870813Z"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Model"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\adamw\\PycharmProjects\\pythonProject\\venv\\lib\\site-packages\\keras\\src\\layers\\core\\dense.py:86: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Dense(256, input_dim=X_train.shape[1], activation='relu', kernel_regularizer=l2(0.001)))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(128, activation='relu', kernel_regularizer=l2(0.001)))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(1, activation='sigmoid'))"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:22:03.872859Z",
     "end_time": "2024-05-19T18:22:04.122687Z"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m6s\u001B[0m 2ms/step - accuracy: 0.8769 - loss: 0.4540 - val_accuracy: 0.9310 - val_loss: 0.2222\n",
      "Epoch 2/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9270 - loss: 0.2362 - val_accuracy: 0.9303 - val_loss: 0.2106\n",
      "Epoch 3/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9320 - loss: 0.2191 - val_accuracy: 0.9415 - val_loss: 0.1890\n",
      "Epoch 4/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9306 - loss: 0.2139 - val_accuracy: 0.9406 - val_loss: 0.1850\n",
      "Epoch 5/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9322 - loss: 0.2098 - val_accuracy: 0.9395 - val_loss: 0.1883\n",
      "Epoch 6/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9325 - loss: 0.2074 - val_accuracy: 0.9404 - val_loss: 0.1814\n",
      "Epoch 7/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9320 - loss: 0.2093 - val_accuracy: 0.9441 - val_loss: 0.1810\n",
      "Epoch 8/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9326 - loss: 0.2094 - val_accuracy: 0.9441 - val_loss: 0.1804\n",
      "Epoch 9/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9327 - loss: 0.2064 - val_accuracy: 0.9400 - val_loss: 0.1807\n",
      "Epoch 10/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9319 - loss: 0.2073 - val_accuracy: 0.9408 - val_loss: 0.1799\n",
      "Epoch 11/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9324 - loss: 0.2061 - val_accuracy: 0.9391 - val_loss: 0.1826\n",
      "Epoch 12/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9320 - loss: 0.2066 - val_accuracy: 0.9433 - val_loss: 0.1814\n",
      "Epoch 13/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9325 - loss: 0.2066 - val_accuracy: 0.9382 - val_loss: 0.1882\n",
      "Epoch 14/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9330 - loss: 0.2045 - val_accuracy: 0.9406 - val_loss: 0.1813\n",
      "Epoch 15/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9316 - loss: 0.2106 - val_accuracy: 0.9408 - val_loss: 0.1831\n",
      "Epoch 16/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9338 - loss: 0.2036 - val_accuracy: 0.9384 - val_loss: 0.1862\n",
      "Epoch 17/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9330 - loss: 0.2063 - val_accuracy: 0.9398 - val_loss: 0.1862\n",
      "Epoch 18/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9320 - loss: 0.2102 - val_accuracy: 0.9408 - val_loss: 0.1802\n",
      "Epoch 19/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9323 - loss: 0.2059 - val_accuracy: 0.9397 - val_loss: 0.1794\n",
      "Epoch 20/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9338 - loss: 0.2039 - val_accuracy: 0.9431 - val_loss: 0.1728\n",
      "Epoch 21/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9319 - loss: 0.2102 - val_accuracy: 0.9415 - val_loss: 0.1787\n",
      "Epoch 22/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9351 - loss: 0.2034 - val_accuracy: 0.9433 - val_loss: 0.1780\n",
      "Epoch 23/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9330 - loss: 0.2059 - val_accuracy: 0.9404 - val_loss: 0.1759\n",
      "Epoch 24/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 2ms/step - accuracy: 0.9335 - loss: 0.2042 - val_accuracy: 0.9409 - val_loss: 0.1789\n",
      "Epoch 25/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9341 - loss: 0.2052 - val_accuracy: 0.9389 - val_loss: 0.1813\n",
      "Epoch 26/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9322 - loss: 0.2078 - val_accuracy: 0.9406 - val_loss: 0.1813\n",
      "Epoch 27/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9319 - loss: 0.2069 - val_accuracy: 0.9283 - val_loss: 0.2017\n",
      "Epoch 28/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9324 - loss: 0.2083 - val_accuracy: 0.9409 - val_loss: 0.1883\n",
      "Epoch 29/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9326 - loss: 0.2054 - val_accuracy: 0.9411 - val_loss: 0.1791\n",
      "Epoch 30/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9333 - loss: 0.2041 - val_accuracy: 0.9419 - val_loss: 0.1769\n",
      "Epoch 31/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9343 - loss: 0.2029 - val_accuracy: 0.9439 - val_loss: 0.1756\n",
      "Epoch 32/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m5s\u001B[0m 1ms/step - accuracy: 0.9330 - loss: 0.2060 - val_accuracy: 0.9384 - val_loss: 0.1805\n",
      "Epoch 33/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9333 - loss: 0.2023 - val_accuracy: 0.9395 - val_loss: 0.1780\n",
      "Epoch 34/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9347 - loss: 0.2025 - val_accuracy: 0.9408 - val_loss: 0.1806\n",
      "Epoch 35/35\n",
      "\u001B[1m3067/3067\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m4s\u001B[0m 1ms/step - accuracy: 0.9315 - loss: 0.2038 - val_accuracy: 0.9419 - val_loss: 0.1762\n"
     ]
    },
    {
     "data": {
      "text/plain": "<keras.src.callbacks.history.History at 0x280fa5dcb80>"
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "model.fit(X_train, y_train, epochs=35, batch_size=32, validation_data=(X_dev, dev_labels))"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:22:04.124694Z",
     "end_time": "2024-05-19T18:24:44.659379Z"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Ewaluacja modelu na zbiorze walidacyjnym"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001B[1m171/171\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 814us/step - accuracy: 0.9413 - loss: 0.1863\n",
      "Accuracy on validation set: 0.9418562054634094\n"
     ]
    }
   ],
   "source": [
    "loss, accuracy = model.evaluate(X_dev, dev_labels)\n",
    "print(f'Accuracy on validation set: {accuracy}')"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:24:44.661382Z",
     "end_time": "2024-05-19T18:24:44.864668Z"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Predykcja na danych walidacyjnych oraz testowych"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001B[1m171/171\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 900us/step\n",
      "\u001B[1m171/171\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 765us/step\n"
     ]
    }
   ],
   "source": [
    "dev_predictions = model.predict(X_dev)\n",
    "test_predictions = model.predict(X_test)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:24:44.863671Z",
     "end_time": "2024-05-19T18:24:45.395043Z"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Zapis wyników"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "outputs": [],
   "source": [
    "dev_predictions = (dev_predictions > 0.5).astype(int)\n",
    "test_predictions = (test_predictions > 0.5).astype(int)\n",
    "\n",
    "pd.DataFrame(dev_predictions).to_csv('dev-0/out.tsv', sep='\\t', header=False, index=False)\n",
    "pd.DataFrame(test_predictions).to_csv('test-A/out.tsv', sep='\\t', header=False, index=False)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "start_time": "2024-05-19T18:24:45.398007Z",
     "end_time": "2024-05-19T18:24:45.438575Z"
    }
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}