challenge_2016_rnn.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Recurrent neural network regression\n",
    "\n",
    "This notebook was used by team \"OpenBikes\" (consisting solely of Max Halford) during the OpenBikes Challenge in 2016. \n",
    "\n",
    "Initially I simply calculated the average number of bikes per hour and per day at each station. This somewhat naïve method - if it even deserves being called that - nonetheless did quite well, thus I wanted to stick to it and improve on top of it. I also wanted to try out \"deep learning\" so I decided to look into recurrent neural networks for regression. My goal was to forecast the first 4 moments for each station with the neural network and see if that made any improvement.\n",
    "\n",
    "A recurrent neural network requires as input a tensor - which is nothing more than a 3D matrix. The tensor has dimensions `(n, p, k)` where\n",
    "\n",
    "- `n` is the number of sequences\n",
    "- `p` is the size of each sequence\n",
    "- `k` the number of values at each point of the sequences\n",
    "\n",
    "The thing is recurrent neural networks work on sequences of observations and try to infer a link between the consecutive values of the sequences. This is easy to understand if you think about a machine that has to classify what people are doing in a movie frame: without knowing what happened in the previous frames than he is going to seriously struggle. The machine needs \"context\" to be able to make a good decision. In our case it's pretty much the same: it's easier to predict how many bikes will be at a station if we know how many there were during the last hour - one of the reasons is because we'll be able to observe an upward/downward trend. `p` will be the number of past observations to look at, `k` will be the number of features to look at each hour - for example the number of bikes, the temperature etc. - and `n` will be the number of sequences that we will feed to the neural network.\n",
    "\n",
    "Basically what I did was to resample the data to obtain a sequence of equidistant observations. In the end I went with 10 minutes because it involved less observations - and less computation! - but also because that was the gap used for the submissions. Thus if I set `p` to 20 then I would be looking back at the $10 \\times 10 = 200$ previous minutes.\n",
    "\n",
    "One tricky thing to deal with is the architecture of the neural network - it's as much an art than a science. However I found that using two LSTM networks seperated by two dropouts - with a 0.2 dropout ratio - worked fine."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data loading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "\n",
    "def get_station_updates(city, station):\n",
    "    df = pd.read_csv('challenge_data/part_1/{}/stations/{}.csv'.format(city, station))\n",
    "    # We're predicting the number of bikes, not the number of spaces\n",
    "    df.drop('spaces', axis=1, inplace=True)\n",
    "    # Maybe some rows are duplicates, who knows...\n",
    "    df.drop_duplicates(subset='moment', inplace=True)\n",
    "    # Index by moment\n",
    "    df['moment'] = pd.to_datetime(df['moment'])\n",
    "    df.set_index('moment', drop=True, inplace=True)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pre-processing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "\n",
    "\n",
    "def add_temporal_features(df):\n",
    "    df['weekday'] = df.index.map(lambda m: m.weekday())\n",
    "    df['hour'] = df.index.map(lambda m: m.hour)\n",
    "    df['minute'] = df.index.map(lambda m: m.minute)\n",
    "    return df\n",
    "\n",
    "\n",
    "def preprocess(df):\n",
    "    # Resample every 10 minutes and forward fill in the missing rows\n",
    "    df = df.resample('10min', label='right', closed='right').ffill()\n",
    "    # Extract temporal features\n",
    "    df = add_temporal_features(df)\n",
    "    # Floating point numbers are more suited for neural nets\n",
    "    df = df.astype('float32')\n",
    "    return df.values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reshape into mini-timeseries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def reshape(data, look_back=1):\n",
    "    X, y = [], []\n",
    "    for i in range(look_back, len(data)):\n",
    "        X.append(data[i-look_back:i])\n",
    "        y.append(data[i][0])\n",
    "    X = np.array(X)\n",
    "    y = np.array(y)\n",
    "    return X, y"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train/test split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For a normal regression problem we could use cross-validation to evaluate our model. With time series data the sequence of values is important, we have to take into account the order of the data. Predicting the past with current data doesn't make any sense."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def train_test_split(X, y, train_ratio=0.8):\n",
    "    train_size = int(len(X) * train_ratio)\n",
    "    test_size = len(X) - train_size\n",
    "    X_train, X_test = X[:train_size], X[-test_size:]\n",
    "    y_train, y_test = y[:train_size], y[-test_size:]\n",
    "    return X_train, X_test, y_train, y_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## RNN on a single station"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Number of previous observations to take into account for forecasting\n",
    "look_back = 12\n",
    "\n",
    "df = get_station_updates('toulouse', '00003-pomme') # pandas dataframe\n",
    "data = preprocess(df) # numpy nd-array\n",
    "X, y = reshape(data, look_back=look_back)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X[-20000:], y[-20000:])\n",
    "\n",
    "n_features = len(X_train[0, 0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "16000/16000 [==============================] - 30s - loss: 5.2082    \n",
      "Epoch 2/5\n",
      "16000/16000 [==============================] - 25s - loss: 3.7752    \n",
      "Epoch 3/5\n",
      "16000/16000 [==============================] - 26s - loss: 2.7051    \n",
      "Epoch 4/5\n",
      "16000/16000 [==============================] - 25s - loss: 2.2234    \n",
      "Epoch 5/5\n",
      "16000/16000 [==============================] - 27s - loss: 2.0538    \n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1117abb38>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from keras import models\n",
    "from keras import layers\n",
    "\n",
    "\n",
    "model = models.Sequential()\n",
    "topology = [n_features, 10, 10, 1]\n",
    "\n",
    "model.add(layers.LSTM(\n",
    "    input_dim=topology[0],\n",
    "    output_dim=topology[1],\n",
    "    input_length=look_back,\n",
    "    return_sequences=True,\n",
    "    unroll=True\n",
    "))\n",
    "model.add(layers.Dropout(0.2))\n",
    "model.add(layers.LSTM(topology[2], return_sequences=False))\n",
    "model.add(layers.Dropout(0.2))\n",
    "model.add(layers.Dense(output_dim=topology[3]))\n",
    "model.add(layers.Activation('linear'))\n",
    "model.compile(loss='mean_absolute_error', optimizer='sgd')\n",
    "model.fit(X_train, y_train, nb_epoch=5, batch_size=64, verbose=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can then call the `.evaluate()` method of the neural network to predict all the values in the test set and compare these predictions to the true values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test mean absolute error: 1.63\n"
     ]
    }
   ],
   "source": [
    "# Estimate model performance\n",
    "test_score = model.evaluate(X_test, y_test, verbose=0)\n",
    "print('Test mean absolute error: {:.2f}'.format(test_score))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also plot some of the predicted values against the true values from the test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x115b0ab00>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6wAAAKvCAYAAABu2EJeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11279cb38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline\n",
    "plt.style.use('ggplot')\n",
    "\n",
    "# Generate predictions for training\n",
    "y_pred = model.predict(X_test)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(16,12))\n",
    "ax.plot(y_test[-1000:], alpha=0.5, label='Truth')\n",
    "ax.plot(y_pred[-1000:], alpha=0.5, label='Predicted')\n",
    "ax.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This works but it is deceptive. At every time step $t$ the last $n-1$ observations that were used for predicting were known to be true. This is fine if we want to predict the next time step, however in our case we will have to predict multiple points in the future.\n",
    "\n",
    "To do so we will have to simulate the time steps one-by-one by using a sliding window. We set an initial window to contain $k$ observations. We use these $k$ observations to predict the next obsersation. We then \"slide\" the window by removing the first observation and putting the predicted observation at the end of the window. After $k$ steps none of the initial observations will remain because they will have been replaced with predicted observations one-by-one.\n",
    "\n",
    "This and other matters are illustrated in one of [Jakob Aungiers's blog posts](http://www.jakob-aungiers.com/articles/a/LSTM-Neural-Network-for-Time-Series-Prediction)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train an RNN for each station in the test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 - 00003-pomme - 24.81 s.\n",
      "1 - 00040-colombette - 47.67 s.\n",
      "2 - 00055-st-sernin-g-arnoult - 70.47 s.\n",
      "3 - 00065-verdier-grand-rond - 93.68 s.\n",
      "4 - 00161-rangueil-bedouce - 116.74 s.\n",
      "5 - 00189-muret-boutinon - 140.48 s.\n",
      "6 - 00204-pl-roger-arnaud - 165.26 s.\n",
      "7 - 00210-terrasse-centre-co - 188.58 s.\n",
      "8 - 00229-iut-rangueil - 212.93 s.\n",
      "9 - 00243-etats-unis-ferry - 237.77 s.\n",
      "10 - 01020-rivoli-concorde - 261.79 s.\n",
      "11 - 04013-francs-bourgeois - 286.93 s.\n",
      "12 - 07008-solferino - 310.70 s.\n",
      "13 - 08035-narvick - 333.31 s.\n",
      "14 - 08046-alma-marceau - 356.87 s.\n",
      "15 - 08101-londres-amsterdam - 381.88 s.\n",
      "16 - 1-ave-e-30-st - 406.70 s.\n",
      "17 - 12029-dugommier - 430.06 s.\n",
      "18 - 15020-mairie-du-15eme - 454.46 s.\n",
      "19 - 23001-couturier-1-levallois - 478.12 s.\n",
      "20 - 3-ave-schermerhorn-st - 504.38 s.\n",
      "21 - 43002-lagny-vincennes - 529.30 s.\n",
      "22 - broad-st-bridge-st - 553.61 s.\n",
      "23 - clinton-st-grand-st - 578.77 s.\n",
      "24 - e-11-st-1-ave - 602.42 s.\n",
      "25 - e-17-st-broadway - 626.58 s.\n",
      "26 - exchange-place - 650.78 s.\n",
      "27 - kent-ave-s-11-st - 673.87 s.\n",
      "28 - pershing-square-north - 696.69 s.\n",
      "29 - w-4-st-7-ave-s - 721.20 s.\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "\n",
    "station_models = {}\n",
    "to_predict_df = pd.read_csv('challenge_data/test-blank.csv', index_col=0)\n",
    "t0 = time.time()\n",
    "\n",
    "for i, (station, group) in enumerate(to_predict_df.groupby('station')):\n",
    "    city = group['city'][0]\n",
    "    df_station = get_station_updates(city, station)\n",
    "    data_station = preprocess(df_station)[-3000:]\n",
    "    X_station, y_station = reshape(data_station, look_back=look_back)\n",
    "    model.fit(X_station, y_station, nb_epoch=5, batch_size=64, verbose=0)\n",
    "    station_models[station] = model\n",
    "    print('{} - {} - {:.2f} s.'.format(i, station, time.time() - t0))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Simulate a scenario for each station in the test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2017-01-04 10:10:00 7.0\n",
      "2017-01-04 10:20:00 6.0\n",
      "2017-01-04 10:30:00 6.0\n",
      "2017-01-04 10:40:00 5.0\n",
      "2017-01-04 10:50:00 4.0\n",
      "2017-01-04 11:00:00 4.0\n",
      "2017-01-04 11:10:00 5.0\n",
      "2017-01-04 11:20:00 4.0\n",
      "2017-01-04 11:30:00 4.0\n",
      "2017-01-04 11:40:00 3.0\n",
      "2017-01-04 11:50:00 3.0\n",
      "2017-01-04 12:00:00 3.0\n",
      "0 - 00003-pomme - 0.52 s.\n",
      "2017-01-04 10:10:00 4.0\n",
      "2017-01-04 10:20:00 4.0\n",
      "2017-01-04 10:30:00 3.0\n",
      "2017-01-04 10:40:00 3.0\n",
      "2017-01-04 10:50:00 2.0\n",
      "2017-01-04 11:00:00 2.0\n",
      "2017-01-04 11:10:00 2.0\n",
      "2017-01-04 11:20:00 2.0\n",
      "2017-01-04 11:30:00 2.0\n",
      "2017-01-04 11:40:00 1.0\n",
      "2017-01-04 11:50:00 1.0\n",
      "2017-01-04 12:00:00 1.0\n",
      "1 - 00040-colombette - 1.01 s.\n",
      "2017-01-04 10:10:00 11.0\n",
      "2017-01-04 10:20:00 10.0\n",
      "2017-01-04 10:30:00 9.0\n",
      "2017-01-04 10:40:00 8.0\n",
      "2017-01-04 10:50:00 7.0\n",
      "2017-01-04 11:00:00 8.0\n",
      "2017-01-04 11:10:00 8.0\n",
      "2017-01-04 11:20:00 8.0\n",
      "2017-01-04 11:30:00 8.0\n",
      "2017-01-04 11:40:00 7.0\n",
      "2017-01-04 11:50:00 6.0\n",
      "2017-01-04 12:00:00 7.0\n",
      "2 - 00055-st-sernin-g-arnoult - 1.45 s.\n",
      "2017-01-04 10:10:00 20.0\n",
      "2017-01-04 10:20:00 21.0\n",
      "2017-01-04 10:30:00 19.0\n",
      "2017-01-04 10:40:00 18.0\n",
      "2017-01-04 10:50:00 17.0\n",
      "2017-01-04 11:00:00 18.0\n",
      "2017-01-04 11:10:00 19.0\n",
      "2017-01-04 11:20:00 20.0\n",
      "2017-01-04 11:30:00 19.0\n",
      "2017-01-04 11:40:00 18.0\n",
      "2017-01-04 11:50:00 17.0\n",
      "2017-01-04 12:00:00 19.0\n",
      "3 - 00065-verdier-grand-rond - 1.99 s.\n",
      "2017-01-04 10:10:00 7.0\n",
      "2017-01-04 10:20:00 7.0\n",
      "2017-01-04 10:30:00 6.0\n",
      "2017-01-04 10:40:00 6.0\n",
      "2017-01-04 10:50:00 5.0\n",
      "2017-01-04 11:00:00 5.0\n",
      "2017-01-04 11:10:00 6.0\n",
      "2017-01-04 11:20:00 6.0\n",
      "2017-01-04 11:30:00 6.0\n",
      "2017-01-04 11:40:00 5.0\n",
      "2017-01-04 11:50:00 4.0\n",
      "2017-01-04 12:00:00 4.0\n",
      "4 - 00161-rangueil-bedouce - 2.47 s.\n",
      "2017-01-04 10:10:00 1.0\n",
      "2017-01-04 10:20:00 1.0\n",
      "2017-01-04 10:30:00 1.0\n",
      "2017-01-04 10:40:00 1.0\n",
      "2017-01-04 10:50:00 1.0\n",
      "2017-01-04 11:00:00 1.0\n",
      "2017-01-04 11:10:00 1.0\n",
      "2017-01-04 11:20:00 1.0\n",
      "2017-01-04 11:30:00 1.0\n",
      "2017-01-04 11:40:00 1.0\n",
      "2017-01-04 11:50:00 1.0\n",
      "2017-01-04 12:00:00 1.0\n",
      "5 - 00189-muret-boutinon - 3.14 s.\n",
      "2017-01-04 10:10:00 2.0\n",
      "2017-01-04 10:20:00 2.0\n",
      "2017-01-04 10:30:00 2.0\n",
      "2017-01-04 10:40:00 1.0\n",
      "2017-01-04 10:50:00 1.0\n",
      "2017-01-04 11:00:00 1.0\n",
      "2017-01-04 11:10:00 1.0\n",
      "2017-01-04 11:20:00 1.0\n",
      "2017-01-04 11:30:00 1.0\n",
      "2017-01-04 11:40:00 1.0\n",
      "2017-01-04 11:50:00 1.0\n",
      "2017-01-04 12:00:00 1.0\n",
      "6 - 00204-pl-roger-arnaud - 3.57 s.\n",
      "2017-01-04 10:10:00 6.0\n",
      "2017-01-04 10:20:00 5.0\n",
      "2017-01-04 10:30:00 4.0\n",
      "2017-01-04 10:40:00 3.0\n",
      "2017-01-04 10:50:00 3.0\n",
      "2017-01-04 11:00:00 3.0\n",
      "2017-01-04 11:10:00 3.0\n",
      "2017-01-04 11:20:00 3.0\n",
      "2017-01-04 11:30:00 2.0\n",
      "2017-01-04 11:40:00 2.0\n",
      "2017-01-04 11:50:00 2.0\n",
      "2017-01-04 12:00:00 2.0\n",
      "7 - 00210-terrasse-centre-co - 4.04 s.\n",
      "2017-01-04 10:10:00 5.0\n",
      "2017-01-04 10:20:00 5.0\n",
      "2017-01-04 10:30:00 4.0\n",
      "2017-01-04 10:40:00 4.0\n",
      "2017-01-04 10:50:00 3.0\n",
      "2017-01-04 11:00:00 3.0\n",
      "2017-01-04 11:10:00 3.0\n",
      "2017-01-04 11:20:00 3.0\n",
      "2017-01-04 11:30:00 2.0\n",
      "2017-01-04 11:40:00 2.0\n",
      "2017-01-04 11:50:00 2.0\n",
      "2017-01-04 12:00:00 2.0\n",
      "8 - 00229-iut-rangueil - 4.50 s.\n",
      "2017-01-04 10:10:00 6.0\n",
      "2017-01-04 10:20:00 6.0\n",
      "2017-01-04 10:30:00 5.0\n",
      "2017-01-04 10:40:00 5.0\n",
      "2017-01-04 10:50:00 4.0\n",
      "2017-01-04 11:00:00 4.0\n",
      "2017-01-04 11:10:00 4.0\n",
      "2017-01-04 11:20:00 4.0\n",
      "2017-01-04 11:30:00 4.0\n",
      "2017-01-04 11:40:00 3.0\n",
      "2017-01-04 11:50:00 3.0\n",
      "2017-01-04 12:00:00 3.0\n",
      "9 - 00243-etats-unis-ferry - 4.99 s.\n",
      "2017-01-04 10:10:00 12.0\n",
      "2017-01-04 10:20:00 12.0\n",
      "2017-01-04 10:30:00 12.0\n",
      "2017-01-04 10:40:00 10.0\n",
      "2017-01-04 10:50:00 9.0\n",
      "2017-01-04 11:00:00 9.0\n",
      "2017-01-04 11:10:00 11.0\n",
      "2017-01-04 11:20:00 10.0\n",
      "2017-01-04 11:30:00 11.0\n",
      "2017-01-04 11:40:00 8.0\n",
      "2017-01-04 11:50:00 7.0\n",
      "2017-01-04 12:00:00 7.0\n",
      "10 - 01020-rivoli-concorde - 5.54 s.\n",
      "2017-01-04 10:10:00 1.0\n",
      "2017-01-04 10:20:00 1.0\n",
      "2017-01-04 10:30:00 1.0\n",
      "2017-01-04 10:40:00 1.0\n",
      "2017-01-04 10:50:00 1.0\n",
      "2017-01-04 11:00:00 1.0\n",
      "2017-01-04 11:10:00 1.0\n",
      "2017-01-04 11:20:00 1.0\n",
      "2017-01-04 11:30:00 1.0\n",
      "2017-01-04 11:40:00 1.0\n",
      "2017-01-04 11:50:00 1.0\n",
      "2017-01-04 12:00:00 1.0\n",
      "11 - 04013-francs-bourgeois - 5.97 s.\n",
      "2017-01-04 10:10:00 8.0\n",
      "2017-01-04 10:20:00 7.0\n",
      "2017-01-04 10:30:00 7.0\n",
      "2017-01-04 10:40:00 6.0\n",
      "2017-01-04 10:50:00 5.0\n",
      "2017-01-04 11:00:00 6.0\n",
      "2017-01-04 11:10:00 6.0\n",
      "2017-01-04 11:20:00 6.0\n",
      "2017-01-04 11:30:00 6.0\n",
      "2017-01-04 11:40:00 5.0\n",
      "2017-01-04 11:50:00 4.0\n",
      "2017-01-04 12:00:00 4.0\n",
      "12 - 07008-solferino - 6.35 s.\n",
      "2017-01-04 10:10:00 3.0\n",
      "2017-01-04 10:20:00 3.0\n",
      "2017-01-04 10:30:00 3.0\n",
      "2017-01-04 10:40:00 2.0\n",
      "2017-01-04 10:50:00 2.0\n",
      "2017-01-04 11:00:00 2.0\n",
      "2017-01-04 11:10:00 2.0\n",
      "2017-01-04 11:20:00 2.0\n",
      "2017-01-04 11:30:00 2.0\n",
      "2017-01-04 11:40:00 1.0\n",
      "2017-01-04 11:50:00 1.0\n",
      "2017-01-04 12:00:00 1.0\n",
      "13 - 08035-narvick - 6.75 s.\n",
      "2017-01-04 10:10:00 14.0\n",
      "2017-01-04 10:20:00 13.0\n",
      "2017-01-04 10:30:00 13.0\n",
      "2017-01-04 10:40:00 11.0\n",
      "2017-01-04 10:50:00 10.0\n",
      "2017-01-04 11:00:00 12.0\n",
      "2017-01-04 11:10:00 13.0\n",
      "2017-01-04 11:20:00 12.0\n",
      "2017-01-04 11:30:00 13.0\n",
      "2017-01-04 11:40:00 12.0\n",
      "2017-01-04 11:50:00 10.0\n",
      "2017-01-04 12:00:00 11.0\n",
      "14 - 08046-alma-marceau - 7.15 s.\n",
      "2017-01-04 10:10:00 7.0\n",
      "2017-01-04 10:20:00 6.0\n",
      "2017-01-04 10:30:00 5.0\n",
      "2017-01-04 10:40:00 4.0\n",
      "2017-01-04 10:50:00 3.0\n",
      "2017-01-04 11:00:00 3.0\n",
      "2017-01-04 11:10:00 3.0\n",
      "2017-01-04 11:20:00 3.0\n",
      "2017-01-04 11:30:00 2.0\n",
      "2017-01-04 11:40:00 2.0\n",
      "2017-01-04 11:50:00 2.0\n",
      "2017-01-04 12:00:00 2.0\n",
      "15 - 08101-londres-amsterdam - 7.69 s.\n",
      "2017-01-04 10:10:00 24.0\n",
      "2017-01-04 10:20:00 25.0\n",
      "2017-01-04 10:30:00 23.0\n",
      "2017-01-04 10:40:00 21.0\n",
      "2017-01-04 10:50:00 19.0\n",
      "2017-01-04 11:00:00 19.0\n",
      "2017-01-04 11:10:00 22.0\n",
      "2017-01-04 11:20:00 23.0\n",
      "2017-01-04 11:30:00 23.0\n",
      "2017-01-04 11:40:00 21.0\n",
      "2017-01-04 11:50:00 20.0\n",
      "2017-01-04 12:00:00 20.0\n",
      "16 - 1-ave-e-30-st - 8.13 s.\n",
      "2017-01-04 10:10:00 23.0\n",
      "2017-01-04 10:20:00 26.0\n",
      "2017-01-04 10:30:00 26.0\n",
      "2017-01-04 10:40:00 24.0\n",
      "2017-01-04 10:50:00 22.0\n",
      "2017-01-04 11:00:00 21.0\n",
      "2017-01-04 11:10:00 22.0\n",
      "2017-01-04 11:20:00 24.0\n",
      "2017-01-04 11:30:00 26.0\n",
      "2017-01-04 11:40:00 25.0\n",
      "2017-01-04 11:50:00 23.0\n",
      "2017-01-04 12:00:00 22.0\n",
      "17 - 12029-dugommier - 8.53 s.\n",
      "2017-01-04 10:10:00 8.0\n",
      "2017-01-04 10:20:00 8.0\n",
      "2017-01-04 10:30:00 7.0\n",
      "2017-01-04 10:40:00 6.0\n",
      "2017-01-04 10:50:00 5.0\n",
      "2017-01-04 11:00:00 5.0\n",
      "2017-01-04 11:10:00 6.0\n",
      "2017-01-04 11:20:00 6.0\n",
      "2017-01-04 11:30:00 6.0\n",
      "2017-01-04 11:40:00 5.0\n",
      "2017-01-04 11:50:00 4.0\n",
      "2017-01-04 12:00:00 4.0\n",
      "18 - 15020-mairie-du-15eme - 8.94 s.\n",
      "2017-01-04 10:10:00 15.0\n",
      "2017-01-04 10:20:00 16.0\n",
      "2017-01-04 10:30:00 17.0\n",
      "2017-01-04 10:40:00 18.0\n",
      "2017-01-04 10:50:00 15.0\n",
      "2017-01-04 11:00:00 16.0\n",
      "2017-01-04 11:10:00 17.0\n",
      "2017-01-04 11:20:00 18.0\n",
      "2017-01-04 11:30:00 18.0\n",
      "2017-01-04 11:40:00 19.0\n",
      "2017-01-04 11:50:00 18.0\n",
      "2017-01-04 12:00:00 18.0\n",
      "19 - 23001-couturier-1-levallois - 9.34 s.\n",
      "2017-01-04 10:10:00 11.0\n",
      "2017-01-04 10:20:00 10.0\n",
      "2017-01-04 10:30:00 9.0\n",
      "2017-01-04 10:40:00 8.0\n",
      "2017-01-04 10:50:00 7.0\n",
      "2017-01-04 11:00:00 8.0\n",
      "2017-01-04 11:10:00 9.0\n",
      "2017-01-04 11:20:00 9.0\n",
      "2017-01-04 11:30:00 8.0\n",
      "2017-01-04 11:40:00 7.0\n",
      "2017-01-04 11:50:00 7.0\n",
      "2017-01-04 12:00:00 7.0\n",
      "20 - 3-ave-schermerhorn-st - 9.99 s.\n",
      "2017-01-04 10:10:00 12.0\n",
      "2017-01-04 10:20:00 13.0\n",
      "2017-01-04 10:30:00 15.0\n",
      "2017-01-04 10:40:00 13.0\n",
      "2017-01-04 10:50:00 10.0\n",
      "2017-01-04 11:00:00 11.0\n",
      "2017-01-04 11:10:00 12.0\n",
      "2017-01-04 11:20:00 13.0\n",
      "2017-01-04 11:30:00 15.0\n",
      "2017-01-04 11:40:00 14.0\n",
      "2017-01-04 11:50:00 10.0\n",
      "2017-01-04 12:00:00 12.0\n",
      "21 - 43002-lagny-vincennes - 10.39 s.\n",
      "2017-01-04 10:10:00 26.0\n",
      "2017-01-04 10:20:00 27.0\n",
      "2017-01-04 10:30:00 28.0\n",
      "2017-01-04 10:40:00 27.0\n",
      "2017-01-04 10:50:00 27.0\n",
      "2017-01-04 11:00:00 29.0\n",
      "2017-01-04 11:10:00 29.0\n",
      "2017-01-04 11:20:00 28.0\n",
      "2017-01-04 11:30:00 28.0\n",
      "2017-01-04 11:40:00 26.0\n",
      "2017-01-04 11:50:00 25.0\n",
      "2017-01-04 12:00:00 26.0\n",
      "22 - broad-st-bridge-st - 10.94 s.\n",
      "2017-01-04 10:10:00 1.0\n",
      "2017-01-04 10:20:00 1.0\n",
      "2017-01-04 10:30:00 1.0\n",
      "2017-01-04 10:40:00 1.0\n",
      "2017-01-04 10:50:00 1.0\n",
      "2017-01-04 11:00:00 1.0\n",
      "2017-01-04 11:10:00 1.0\n",
      "2017-01-04 11:20:00 1.0\n",
      "2017-01-04 11:30:00 1.0\n",
      "2017-01-04 11:40:00 1.0\n",
      "2017-01-04 11:50:00 1.0\n",
      "2017-01-04 12:00:00 1.0\n",
      "23 - clinton-st-grand-st - 11.53 s.\n",
      "2017-01-04 10:10:00 1.0\n",
      "2017-01-04 10:20:00 1.0\n",
      "2017-01-04 10:30:00 1.0\n",
      "2017-01-04 10:40:00 1.0\n",
      "2017-01-04 10:50:00 1.0\n",
      "2017-01-04 11:00:00 1.0\n",
      "2017-01-04 11:10:00 1.0\n",
      "2017-01-04 11:20:00 1.0\n",
      "2017-01-04 11:30:00 1.0\n",
      "2017-01-04 11:40:00 1.0\n",
      "2017-01-04 11:50:00 1.0\n",
      "2017-01-04 12:00:00 1.0\n",
      "24 - e-11-st-1-ave - 12.09 s.\n",
      "2017-01-04 10:10:00 28.0\n",
      "2017-01-04 10:20:00 28.0\n",
      "2017-01-04 10:30:00 29.0\n",
      "2017-01-04 10:40:00 29.0\n",
      "2017-01-04 10:50:00 32.0\n",
      "2017-01-04 11:00:00 32.0\n",
      "2017-01-04 11:10:00 32.0\n",
      "2017-01-04 11:20:00 32.0\n",
      "2017-01-04 11:30:00 33.0\n",
      "2017-01-04 11:40:00 33.0\n",
      "2017-01-04 11:50:00 31.0\n",
      "2017-01-04 12:00:00 30.0\n",
      "25 - e-17-st-broadway - 12.74 s.\n",
      "2017-01-04 10:10:00 22.0\n",
      "2017-01-04 10:20:00 25.0\n",
      "2017-01-04 10:30:00 28.0\n",
      "2017-01-04 10:40:00 26.0\n",
      "2017-01-04 10:50:00 24.0\n",
      "2017-01-04 11:00:00 25.0\n",
      "2017-01-04 11:10:00 26.0\n",
      "2017-01-04 11:20:00 27.0\n",
      "2017-01-04 11:30:00 31.0\n",
      "2017-01-04 11:40:00 30.0\n",
      "2017-01-04 11:50:00 27.0\n",
      "2017-01-04 12:00:00 26.0\n",
      "26 - exchange-place - 13.17 s.\n",
      "2017-01-04 10:10:00 2.0\n",
      "2017-01-04 10:20:00 2.0\n",
      "2017-01-04 10:30:00 2.0\n",
      "2017-01-04 10:40:00 2.0\n",
      "2017-01-04 10:50:00 2.0\n",
      "2017-01-04 11:00:00 2.0\n",
      "2017-01-04 11:10:00 2.0\n",
      "2017-01-04 11:20:00 2.0\n",
      "2017-01-04 11:30:00 2.0\n",
      "2017-01-04 11:40:00 1.0\n",
      "2017-01-04 11:50:00 1.0\n",
      "2017-01-04 12:00:00 1.0\n",
      "27 - kent-ave-s-11-st - 13.66 s.\n",
      "2017-01-04 10:10:00 1.0\n",
      "2017-01-04 10:20:00 1.0\n",
      "2017-01-04 10:30:00 1.0\n",
      "2017-01-04 10:40:00 1.0\n",
      "2017-01-04 10:50:00 1.0\n",
      "2017-01-04 11:00:00 1.0\n",
      "2017-01-04 11:10:00 1.0\n",
      "2017-01-04 11:20:00 1.0\n",
      "2017-01-04 11:30:00 1.0\n",
      "2017-01-04 11:40:00 1.0\n",
      "2017-01-04 11:50:00 1.0\n",
      "2017-01-04 12:00:00 1.0\n",
      "28 - pershing-square-north - 14.12 s.\n",
      "2017-01-04 10:10:00 1.0\n",
      "2017-01-04 10:20:00 1.0\n",
      "2017-01-04 10:30:00 1.0\n",
      "2017-01-04 10:40:00 1.0\n",
      "2017-01-04 10:50:00 1.0\n",
      "2017-01-04 11:00:00 1.0\n",
      "2017-01-04 11:10:00 1.0\n",
      "2017-01-04 11:20:00 1.0\n",
      "2017-01-04 11:30:00 1.0\n",
      "2017-01-04 11:40:00 1.0\n",
      "2017-01-04 11:50:00 1.0\n",
      "2017-01-04 12:00:00 1.0\n",
      "29 - w-4-st-7-ave-s - 14.68 s.\n"
     ]
    }
   ],
   "source": [
    "import datetime as dt\n",
    "\n",
    "\n",
    "scenarios = {}\n",
    "t0 = time.time()\n",
    "\n",
    "for i, (station, group) in enumerate(to_predict_df.groupby('station')):\n",
    "    scenarios[station] = {}\n",
    "    city = group['city'][0]\n",
    "    model = station_models[station]\n",
    "\n",
    "    # Simulate from 10:10 until 12:00 for the first day\n",
    "    since = dt.datetime(year=2017, month=1, day=4, hour=10, minute=10)\n",
    "    until = dt.datetime(year=2017, month=1, day=4, hour=12, minute=0)\n",
    "\n",
    "    # Initialize a sliding window with the most recent observations\n",
    "    df_station = get_station_updates(city, station)\n",
    "    window = preprocess(df_station)[-look_back:].copy()\n",
    "\n",
    "    now = since\n",
    "    while now <= until:\n",
    "        prediction = np.round(model.predict(window.reshape(1, look_back, n_features))[0][0])\n",
    "        print(now, prediction)\n",
    "        scenarios[station][now] = prediction\n",
    "        now += dt.timedelta(minutes=10)\n",
    "        # Update the sliding window\n",
    "        window = np.append(window[1:], [[\n",
    "            prediction,\n",
    "            float(now.weekday()),\n",
    "            float(now.hour),\n",
    "            float(now.minute)\n",
    "        ]], axis=0)\n",
    "\n",
    "    print('{} - {} - {:.2f} s.'.format(i, station, time.time() - t0))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I then use these values and replace the values in the windowed average submission."
   ]
  }
 ],
 "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}