time-series-prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TODOs\n",
    "* Stage data into memory instead of fit_generator\n",
    "* Exercise for 1st part: investigate data, split data in a sensible way, set up a model, train it, evaluate it (simple RNN)\n",
    "* Exercise for 2nd part: LSTMs, GRUs, dropout, recurrent dropout. Stacking? \n",
    "* For assignment: class weights to deal with class imbalance?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "── Attaching packages ─────────────────────────────────────── tidyverse 1.2.1 ──\n",
      "✔ tibble  2.0.1     ✔ purrr   0.2.5\n",
      "✔ tidyr   0.8.2     ✔ dplyr   0.7.8\n",
      "✔ readr   1.3.1     ✔ stringr 1.3.1\n",
      "✔ tibble  2.0.1     ✔ forcats 0.3.0\n",
      "── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──\n",
      "✖ dplyr::filter() masks stats::filter()\n",
      "✖ dplyr::lag()    masks stats::lag()\n"
     ]
    }
   ],
   "source": [
    "library(keras)\n",
    "library(ggplot2)\n",
    "library(tidyverse)\n",
    "library(rsample)\n",
    "library(tictoc)\n",
    "\n",
    "library(repr)\n",
    "options(repr.plot.width=6, repr.plot.height=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df <- read.csv('../erasmus/', stringsAsFactors=FALSE)[,c('Date', 'Open')]\n",
    "df$Date <- as.Date(df$Date, '%Y-%m-%d')\n",
    "head(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "glimpse(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "ggplot(df, aes(df$Date, df$Open)) + geom_line() + xlab('Date') + ylab('Open stock price ($)') + ggtitle('AAPL opening stock price')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing\n",
    "The label of data point `x` should be the opening stock price of data point `x+1`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "?rolling_origin"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tmp <- data.frame(df$Open)\n",
    "samples <- df$Open\n",
    "n <- length(samples)\n",
    "X_train <- array_reshape(head(samples, n - 1), c(n - 1, 1))\n",
    "y_train <- tail(samples, n - 1)\n",
    "cat(dim(X_train), dim(y_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "length(samples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "apply.rolling(tmp, )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "rolling_origin_samples <- rolling_origin(\n",
    "    tmp,\n",
    "    initial=7,\n",
    "    assess=1,\n",
    "    cumulative=FALSE\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dim(rolling_origin_samples$splits[[5]]$data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "glimpse(rolling_origin_samples$splits)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "x <- seq(1, 100, length=1000)\n",
    "y <- sin(x)\n",
    "plot(x, y, type='l')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "generator <- function(x, y, lookback, batch_size) {\n",
    "    len <- length(x)\n",
    "    function() {\n",
    "        # Indices are for the targets\n",
    "        indices <- sample(lookback + 1:len, batch_size)\n",
    "        samples <- array(0, dim())\n",
    "        \n",
    "    }\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dim(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "length(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len <- length(x)\n",
    "lookback <- 7\n",
    "batch_size <- 4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "indices <- sample(lookback + 1:len, batch_size)\n",
    "indices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "samples <- array(0, dim=c())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Build model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model <- keras_model_sequential()\n",
    "model %>%\n",
    "    layer_lstm(\n",
    "        units=128,\n",
    "        input_shape=c(4, 7, 1),\n",
    "        dropout=0.2,\n",
    "        recurrent_dropout=0.2,\n",
    "        return_sequences=TRUE\n",
    "    ) %>% layer_dense(units=1)\n",
    "\n",
    "model %>%\n",
    "    compile(\n",
    "        loss='logcosh',\n",
    "        optimizer=optimizer_sgd(lr=0.1, momentum=0.9),\n",
    "        metrics=list('mean_squared_error')\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "history <- model %>% fit(\n",
    "    x = X_train,\n",
    "    y = y_train,\n",
    "#     validation_data=list(X_val, y_val),\n",
    "    batch_size=4,\n",
    "    epochs=2\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## DEBUG: IMDB sample from book"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_features <- 10000\n",
    "maxlen <- 500\n",
    "batch_size = 64\n",
    "\n",
    "imdb <- dataset_imdb(num_words = max_features)\n",
    "c(c(X_train, y_train), c(X_test, y_test)) %<-% imdb\n",
    "cat(length(X_train))\n",
    "length(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "length(X_train[[1]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train <- pad_sequences(X_train, maxlen = maxlen)\n",
    "X_test  <- pad_sequences(X_test, maxlen = maxlen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dim(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "length(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train[5000,]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model <- keras_model_sequential() %>%\n",
    "    layer_embedding(input_dim = max_features, output_dim = 32) %>%\n",
    "    layer_simple_rnn(units = 32) %>%\n",
    "    layer_dense(units = 1, activation = \"sigmoid\")\n",
    "\n",
    "cat(summary(model))\n",
    "\n",
    "model %>% compile(\n",
    "    optimizer = optimizer_rmsprop(),\n",
    "    loss = \"binary_crossentropy\",\n",
    "    metrics = c(\"acc\")\n",
    ")\n",
    "\n",
    "history <- model %>% fit(\n",
    "    X_train, y_train,\n",
    "    epochs = 10,\n",
    "    batch_size = 512,\n",
    "    validation_split = 0.2\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## DEBUG: Jena weather data example from book"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Observations: 420,551\n",
      "Variables: 15\n",
      "$ Date.Time       <fct> 01.01.2009 00:10:00, 01.01.2009 00:20:00, 01.01.2009 …\n",
      "$ p..mbar.        <dbl> 996.52, 996.57, 996.53, 996.51, 996.51, 996.50, 996.5…\n",
      "$ T..degC.        <dbl> -8.02, -8.41, -8.51, -8.31, -8.27, -8.05, -7.62, -7.6…\n",
      "$ Tpot..K.        <dbl> 265.40, 265.01, 264.91, 265.12, 265.15, 265.38, 265.8…\n",
      "$ Tdew..degC.     <dbl> -8.90, -9.28, -9.31, -9.07, -9.04, -8.78, -8.30, -8.3…\n",
      "$ rh....          <dbl> 93.3, 93.4, 93.9, 94.2, 94.1, 94.4, 94.8, 94.4, 93.8,…\n",
      "$ VPmax..mbar.    <dbl> 3.33, 3.23, 3.21, 3.26, 3.27, 3.33, 3.44, 3.44, 3.36,…\n",
      "$ VPact..mbar.    <dbl> 3.11, 3.02, 3.01, 3.07, 3.08, 3.14, 3.26, 3.25, 3.15,…\n",
      "$ VPdef..mbar.    <dbl> 0.22, 0.21, 0.20, 0.19, 0.19, 0.19, 0.18, 0.19, 0.21,…\n",
      "$ sh..g.kg.       <dbl> 1.94, 1.89, 1.88, 1.92, 1.92, 1.96, 2.04, 2.03, 1.97,…\n",
      "$ H2OC..mmol.mol. <dbl> 3.12, 3.03, 3.02, 3.08, 3.09, 3.15, 3.27, 3.26, 3.16,…\n",
      "$ rho..g.m..3.    <dbl> 1307.75, 1309.80, 1310.24, 1309.19, 1309.00, 1307.86,…\n",
      "$ wv..m.s.        <dbl> 1.03, 0.72, 0.19, 0.34, 0.32, 0.21, 0.18, 0.19, 0.28,…\n",
      "$ max..wv..m.s.   <dbl> 1.75, 1.50, 0.63, 0.50, 0.63, 0.63, 0.63, 0.50, 0.75,…\n",
      "$ wd..deg.        <dbl> 152.3, 136.1, 171.6, 198.0, 214.3, 192.7, 166.5, 118.…\n"
     ]
    }
   ],
   "source": [
    "data <- read.csv('../jena_climate_2009_2016.csv')\n",
    "glimpse(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAHgCAMAAAC7G6qeAAADAFBMVEUAAAABAQECAgIDAwME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\n6OQegZtZAAAAAElFTkSuQmCC",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ggplot(data, aes(x = 1:nrow(data), y = `T..degC.`)) + geom_line()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAHgCAMAAAC7G6qeAAADAFBMVEUAAAABAQECAgIDAwME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\nTkSuQmCC",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ggplot(data[1:1440,], aes(x = 1:1440, y = `T..degC.`)) + geom_line()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<ol class=list-inline>\n",
       "\t<li>420551</li>\n",
       "\t<li>14</li>\n",
       "</ol>\n"
      ],
      "text/latex": [
       "\\begin{enumerate*}\n",
       "\\item 420551\n",
       "\\item 14\n",
       "\\end{enumerate*}\n"
      ],
      "text/markdown": [
       "1. 420551\n",
       "2. 14\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "[1] 420551     14"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data <- data.matrix(data[,-1])\n",
    "# data <- data[,1:4]\n",
    "dim(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train <- data[1:200000,]\n",
    "data <- scale(data, center = apply(X_train, 2, mean), scale = apply(X_train, 2, sd))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train <- data[1:200000,]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>p..mbar.</th><th scope=col>T..degC.</th><th scope=col>Tpot..K.</th><th scope=col>Tdew..degC.</th><th scope=col>rh....</th><th scope=col>VPmax..mbar.</th><th scope=col>VPact..mbar.</th><th scope=col>VPdef..mbar.</th><th scope=col>sh..g.kg.</th><th scope=col>H2OC..mmol.mol.</th><th scope=col>rho..g.m..3.</th><th scope=col>wv..m.s.</th><th scope=col>max..wv..m.s.</th><th scope=col>wd..deg.</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>0.9001452 </td><td>-1.931354 </td><td>-1.982105 </td><td>-1.862796 </td><td>1.072850  </td><td>-1.307418 </td><td>-1.473754 </td><td>-0.7986844</td><td>-1.476264 </td><td>-1.478152 </td><td>2.123745  </td><td>-0.7295027</td><td>-0.7806778</td><td>-0.2761353</td></tr>\n",
       "\t<tr><td>0.9060411 </td><td>-1.975409 </td><td>-2.025665 </td><td>-1.915825 </td><td>1.078828  </td><td>-1.320424 </td><td>-1.495192 </td><td>-0.8007504</td><td>-1.495021 </td><td>-1.499318 </td><td>2.171993  </td><td>-0.9312378</td><td>-0.8879427</td><td>-0.4631733</td></tr>\n",
       "\t<tr><td>0.9013244 </td><td>-1.986705 </td><td>-2.036834 </td><td>-1.920011 </td><td>1.108719  </td><td>-1.323025 </td><td>-1.497574 </td><td>-0.8028163</td><td>-1.498772 </td><td>-1.501669 </td><td>2.182349  </td><td>-1.2761399</td><td>-1.2612245</td><td>-0.0533062</td></tr>\n",
       "\t<tr><td>0.8989660 </td><td>-1.964113 </td><td>-2.013379 </td><td>-1.886519 </td><td>1.126654  </td><td>-1.316522 </td><td>-1.483282 </td><td>-0.8048823</td><td>-1.483767 </td><td>-1.487559 </td><td>2.157636  </td><td>-1.1785261</td><td>-1.3170022</td><td> 0.2514964</td></tr>\n",
       "\t<tr><td>0.8989660 </td><td>-1.959594 </td><td>-2.010028 </td><td>-1.882333 </td><td>1.120676  </td><td>-1.315222 </td><td>-1.480900 </td><td>-0.8048823</td><td>-1.483767 </td><td>-1.485207 </td><td>2.153165  </td><td>-1.1915412</td><td>-1.2612245</td><td> 0.4396888</td></tr>\n",
       "\t<tr><td>0.8977869 </td><td>-1.934742 </td><td>-1.984339 </td><td>-1.846050 </td><td>1.138610  </td><td>-1.307418 </td><td>-1.466608 </td><td>-0.8048823</td><td>-1.468761 </td><td>-1.471096 </td><td>2.126334  </td><td>-1.2631247</td><td>-1.2612245</td><td> 0.1903049</td></tr>\n",
       "\t<tr><td>0.8977869 </td><td>-1.886169 </td><td>-1.936312 </td><td>-1.779065 </td><td>1.162523  </td><td>-1.293113 </td><td>-1.438024 </td><td>-0.8069483</td><td>-1.438749 </td><td>-1.442875 </td><td>2.075027  </td><td>-1.2826474</td><td>-1.2612245</td><td>-0.1121885</td></tr>\n",
       "\t<tr><td>0.8977869 </td><td>-1.886169 </td><td>-1.936312 </td><td>-1.787438 </td><td>1.138610  </td><td>-1.293113 </td><td>-1.440406 </td><td>-0.8048823</td><td>-1.442501 </td><td>-1.445226 </td><td>2.075262  </td><td>-1.2761399</td><td>-1.3170022</td><td>-0.6652204</td></tr>\n",
       "\t<tr><td>0.8977869 </td><td>-1.918928 </td><td>-1.968702 </td><td>-1.839072 </td><td>1.102741  </td><td>-1.303517 </td><td>-1.464226 </td><td>-0.8007504</td><td>-1.465009 </td><td>-1.468744 </td><td>2.110095  </td><td>-1.2175716</td><td>-1.2097373</td><td> 0.1418136</td></tr>\n",
       "\t<tr><td>0.9013244 </td><td>-1.977668 </td><td>-2.027899 </td><td>-1.924198 </td><td>1.060893  </td><td>-1.320424 </td><td>-1.499956 </td><td>-0.7986844</td><td>-1.498772 </td><td>-1.501669 </td><td>2.173170  </td><td>-1.0158364</td><td>-1.1539596</td><td> 0.1014042</td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{llllllllllllll}\n",
       " p..mbar. & T..degC. & Tpot..K. & Tdew..degC. & rh.... & VPmax..mbar. & VPact..mbar. & VPdef..mbar. & sh..g.kg. & H2OC..mmol.mol. & rho..g.m..3. & wv..m.s. & max..wv..m.s. & wd..deg.\\\\\n",
       "\\hline\n",
       "\t 0.9001452  & -1.931354  & -1.982105  & -1.862796  & 1.072850   & -1.307418  & -1.473754  & -0.7986844 & -1.476264  & -1.478152  & 2.123745   & -0.7295027 & -0.7806778 & -0.2761353\\\\\n",
       "\t 0.9060411  & -1.975409  & -2.025665  & -1.915825  & 1.078828   & -1.320424  & -1.495192  & -0.8007504 & -1.495021  & -1.499318  & 2.171993   & -0.9312378 & -0.8879427 & -0.4631733\\\\\n",
       "\t 0.9013244  & -1.986705  & -2.036834  & -1.920011  & 1.108719   & -1.323025  & -1.497574  & -0.8028163 & -1.498772  & -1.501669  & 2.182349   & -1.2761399 & -1.2612245 & -0.0533062\\\\\n",
       "\t 0.8989660  & -1.964113  & -2.013379  & -1.886519  & 1.126654   & -1.316522  & -1.483282  & -0.8048823 & -1.483767  & -1.487559  & 2.157636   & -1.1785261 & -1.3170022 &  0.2514964\\\\\n",
       "\t 0.8989660  & -1.959594  & -2.010028  & -1.882333  & 1.120676   & -1.315222  & -1.480900  & -0.8048823 & -1.483767  & -1.485207  & 2.153165   & -1.1915412 & -1.2612245 &  0.4396888\\\\\n",
       "\t 0.8977869  & -1.934742  & -1.984339  & -1.846050  & 1.138610   & -1.307418  & -1.466608  & -0.8048823 & -1.468761  & -1.471096  & 2.126334   & -1.2631247 & -1.2612245 &  0.1903049\\\\\n",
       "\t 0.8977869  & -1.886169  & -1.936312  & -1.779065  & 1.162523   & -1.293113  & -1.438024  & -0.8069483 & -1.438749  & -1.442875  & 2.075027   & -1.2826474 & -1.2612245 & -0.1121885\\\\\n",
       "\t 0.8977869  & -1.886169  & -1.936312  & -1.787438  & 1.138610   & -1.293113  & -1.440406  & -0.8048823 & -1.442501  & -1.445226  & 2.075262   & -1.2761399 & -1.3170022 & -0.6652204\\\\\n",
       "\t 0.8977869  & -1.918928  & -1.968702  & -1.839072  & 1.102741   & -1.303517  & -1.464226  & -0.8007504 & -1.465009  & -1.468744  & 2.110095   & -1.2175716 & -1.2097373 &  0.1418136\\\\\n",
       "\t 0.9013244  & -1.977668  & -2.027899  & -1.924198  & 1.060893   & -1.320424  & -1.499956  & -0.7986844 & -1.498772  & -1.501669  & 2.173170   & -1.0158364 & -1.1539596 &  0.1014042\\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "p..mbar. | T..degC. | Tpot..K. | Tdew..degC. | rh.... | VPmax..mbar. | VPact..mbar. | VPdef..mbar. | sh..g.kg. | H2OC..mmol.mol. | rho..g.m..3. | wv..m.s. | max..wv..m.s. | wd..deg. | \n",
       "|---|---|---|---|---|---|---|---|---|---|\n",
       "| 0.9001452  | -1.931354  | -1.982105  | -1.862796  | 1.072850   | -1.307418  | -1.473754  | -0.7986844 | -1.476264  | -1.478152  | 2.123745   | -0.7295027 | -0.7806778 | -0.2761353 | \n",
       "| 0.9060411  | -1.975409  | -2.025665  | -1.915825  | 1.078828   | -1.320424  | -1.495192  | -0.8007504 | -1.495021  | -1.499318  | 2.171993   | -0.9312378 | -0.8879427 | -0.4631733 | \n",
       "| 0.9013244  | -1.986705  | -2.036834  | -1.920011  | 1.108719   | -1.323025  | -1.497574  | -0.8028163 | -1.498772  | -1.501669  | 2.182349   | -1.2761399 | -1.2612245 | -0.0533062 | \n",
       "| 0.8989660  | -1.964113  | -2.013379  | -1.886519  | 1.126654   | -1.316522  | -1.483282  | -0.8048823 | -1.483767  | -1.487559  | 2.157636   | -1.1785261 | -1.3170022 |  0.2514964 | \n",
       "| 0.8989660  | -1.959594  | -2.010028  | -1.882333  | 1.120676   | -1.315222  | -1.480900  | -0.8048823 | -1.483767  | -1.485207  | 2.153165   | -1.1915412 | -1.2612245 |  0.4396888 | \n",
       "| 0.8977869  | -1.934742  | -1.984339  | -1.846050  | 1.138610   | -1.307418  | -1.466608  | -0.8048823 | -1.468761  | -1.471096  | 2.126334   | -1.2631247 | -1.2612245 |  0.1903049 | \n",
       "| 0.8977869  | -1.886169  | -1.936312  | -1.779065  | 1.162523   | -1.293113  | -1.438024  | -0.8069483 | -1.438749  | -1.442875  | 2.075027   | -1.2826474 | -1.2612245 | -0.1121885 | \n",
       "| 0.8977869  | -1.886169  | -1.936312  | -1.787438  | 1.138610   | -1.293113  | -1.440406  | -0.8048823 | -1.442501  | -1.445226  | 2.075262   | -1.2761399 | -1.3170022 | -0.6652204 | \n",
       "| 0.8977869  | -1.918928  | -1.968702  | -1.839072  | 1.102741   | -1.303517  | -1.464226  | -0.8007504 | -1.465009  | -1.468744  | 2.110095   | -1.2175716 | -1.2097373 |  0.1418136 | \n",
       "| 0.9013244  | -1.977668  | -2.027899  | -1.924198  | 1.060893   | -1.320424  | -1.499956  | -0.7986844 | -1.498772  | -1.501669  | 2.173170   | -1.0158364 | -1.1539596 |  0.1014042 | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "      p..mbar.  T..degC.  Tpot..K.  Tdew..degC. rh....   VPmax..mbar.\n",
       " [1,] 0.9001452 -1.931354 -1.982105 -1.862796   1.072850 -1.307418   \n",
       " [2,] 0.9060411 -1.975409 -2.025665 -1.915825   1.078828 -1.320424   \n",
       " [3,] 0.9013244 -1.986705 -2.036834 -1.920011   1.108719 -1.323025   \n",
       " [4,] 0.8989660 -1.964113 -2.013379 -1.886519   1.126654 -1.316522   \n",
       " [5,] 0.8989660 -1.959594 -2.010028 -1.882333   1.120676 -1.315222   \n",
       " [6,] 0.8977869 -1.934742 -1.984339 -1.846050   1.138610 -1.307418   \n",
       " [7,] 0.8977869 -1.886169 -1.936312 -1.779065   1.162523 -1.293113   \n",
       " [8,] 0.8977869 -1.886169 -1.936312 -1.787438   1.138610 -1.293113   \n",
       " [9,] 0.8977869 -1.918928 -1.968702 -1.839072   1.102741 -1.303517   \n",
       "[10,] 0.9013244 -1.977668 -2.027899 -1.924198   1.060893 -1.320424   \n",
       "      VPact..mbar. VPdef..mbar. sh..g.kg. H2OC..mmol.mol. rho..g.m..3.\n",
       " [1,] -1.473754    -0.7986844   -1.476264 -1.478152       2.123745    \n",
       " [2,] -1.495192    -0.8007504   -1.495021 -1.499318       2.171993    \n",
       " [3,] -1.497574    -0.8028163   -1.498772 -1.501669       2.182349    \n",
       " [4,] -1.483282    -0.8048823   -1.483767 -1.487559       2.157636    \n",
       " [5,] -1.480900    -0.8048823   -1.483767 -1.485207       2.153165    \n",
       " [6,] -1.466608    -0.8048823   -1.468761 -1.471096       2.126334    \n",
       " [7,] -1.438024    -0.8069483   -1.438749 -1.442875       2.075027    \n",
       " [8,] -1.440406    -0.8048823   -1.442501 -1.445226       2.075262    \n",
       " [9,] -1.464226    -0.8007504   -1.465009 -1.468744       2.110095    \n",
       "[10,] -1.499956    -0.7986844   -1.498772 -1.501669       2.173170    \n",
       "      wv..m.s.   max..wv..m.s. wd..deg.  \n",
       " [1,] -0.7295027 -0.7806778    -0.2761353\n",
       " [2,] -0.9312378 -0.8879427    -0.4631733\n",
       " [3,] -1.2761399 -1.2612245    -0.0533062\n",
       " [4,] -1.1785261 -1.3170022     0.2514964\n",
       " [5,] -1.1915412 -1.2612245     0.4396888\n",
       " [6,] -1.2631247 -1.2612245     0.1903049\n",
       " [7,] -1.2826474 -1.2612245    -0.1121885\n",
       " [8,] -1.2761399 -1.3170022    -0.6652204\n",
       " [9,] -1.2175716 -1.2097373     0.1418136\n",
       "[10,] -1.0158364 -1.1539596     0.1014042"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train[1:10,]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "generator <- function(data, lookback, delay, min_index, max_index, shuffle = FALSE, batch_size = 512, step = 6) {\n",
    "    if (is.null(max_index))\n",
    "        max_index <- nrow(data) - delay - 1\n",
    "    i <- min_index + lookback  # for looping through the data from start to end and back again\n",
    "    \n",
    "    function() {\n",
    "        if (shuffle)\n",
    "            rows <- sample(c((min_index + lookback):max_index), size = batch_size)\n",
    "        else {\n",
    "            # Loop around to start if necessary\n",
    "            if (i + batch_size >= max_index)\n",
    "                i <<- min_index + lookback\n",
    "            \n",
    "            # Select next batch_size rows\n",
    "            rows <- c(i:min(i + batch_size, max_index))\n",
    "            i <<- i + length(rows)\n",
    "        }\n",
    "        \n",
    "        samples <- array(0, dim = c(length(rows), lookback / step, dim(data)[[-1]]))\n",
    "        targets <- array(0, dim = c(length(rows)))\n",
    "        \n",
    "        for (j in 1:length(rows)) {\n",
    "            indices <- seq(rows[[j]] - lookback, rows[[j]], length.out = dim(samples)[[2]])\n",
    "            samples[j,,] <- data[indices,]\n",
    "            targets[[j]] <- data[rows[[j]] + delay, 2]\n",
    "        }\n",
    "        list(samples, targets)\n",
    "    }\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# lookback <- 7 * 6 * 24\n",
    "lookback <- 1440\n",
    "step <- 6\n",
    "delay <- 144\n",
    "batch_size <- 128\n",
    "steps_per_epoch = 250\n",
    "num_epochs = 25\n",
    "\n",
    "training_generator <- generator(\n",
    "    data,\n",
    "    lookback = lookback,\n",
    "    delay = delay,\n",
    "    min_index = 1,\n",
    "    max_index = 250000,\n",
    "    shuffle = TRUE,\n",
    "    step = step,\n",
    "    batch_size = batch_size\n",
    ")\n",
    "\n",
    "validation_generator <- generator(\n",
    "    data,\n",
    "    lookback = lookback,\n",
    "    delay = delay,\n",
    "    min_index = 250001,\n",
    "    max_index = 300000,\n",
    "    step = step,\n",
    "    batch_size = batch_size\n",
    ")\n",
    "\n",
    "validation_steps = (300000 - 250001 - lookback) / batch_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "Progress <- R6::R6Class(\"Progress\",\n",
    "  inherit = KerasCallback,\n",
    "  \n",
    "  public = list(\n",
    "    num_batches = NULL,\n",
    "    num_epochs  = NULL,\n",
    "    update_frequency = NULL,\n",
    "    epoch = NULL,\n",
    "    batch = NULL,\n",
    "      \n",
    "    initialize = function(num_batches, num_epochs, update_frequency) {\n",
    "        self$num_batches <- num_batches\n",
    "        self$num_epochs <- num_epochs\n",
    "        self$update_frequency <- update_frequency\n",
    "        self$epoch <- 1\n",
    "    },\n",
    "                        \n",
    "    on_batch_end = function(batch, logs = list()) {\n",
    "        if ((batch + 1) %% self$update_frequency == 0) {\n",
    "            cat('Epoch ', self$epoch + 1, '/', self$num_epochs, ': ', batch + 1, '/', self$num_batches, \n",
    "                ' - loss: ', logs[['loss']], ', acc.: ', logs[['acc']], '\\r', sep = '')\n",
    "            flush.console()\n",
    "        }\n",
    "        self$batch <- batch\n",
    "    },\n",
    "      \n",
    "    on_epoch_begin = function(epoch, logs = list()) {\n",
    "        self$epoch <- epoch\n",
    "    },\n",
    "                        \n",
    "    on_epoch_end = function(epoch, logs = list()) {\n",
    "        cat('Epoch ', self$epoch + 1, '/', self$num_epochs, ': ', self$batch + 1, '/', self$num_batches, \n",
    "                ' - loss: ', logs[['loss']], ', acc.: ',  logs[['acc']], ' - val. loss: ', logs[['val_loss']],\n",
    "                ' - val. acc.: ', logs[['val_acc']], '\\n', sep = '')\n",
    "        flush.console()\n",
    "    }\n",
    "))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model <- keras_model_sequential() %>%\n",
    "#     layer_simple_rnn(units = 32, input_shape = list(NULL, dim(data)[[-1]])) %>%\n",
    "#     layer_gru(units = 32, input_shape = list(NULL, dim(data)[[-1]])) %>%\n",
    "    layer_gru(units = 32, input_shape = list(NULL, dim(data)[[-1]]), recurrent_dropout = 0.2, dropout = 0.2) %>%\n",
    "#     layer_lstm(units = 32) %>%\n",
    "    layer_dense(units = 1)\n",
    "cat(summary(model))\n",
    "\n",
    "progress <- Progress$new(steps_per_epoch, num_epochs, 10)\n",
    "\n",
    "model %>% compile(\n",
    "    optimizer = optimizer_rmsprop(),\n",
    "    loss = \"mae\"\n",
    ")\n",
    "\n",
    "history <- model %>% fit_generator(\n",
    "    training_generator,\n",
    "    steps_per_epoch = steps_per_epoch,\n",
    "    epochs = num_epochs,\n",
    "    validation_data = validation_generator,\n",
    "    validation_steps = validation_steps,\n",
    "    callbacks = list(progress)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Regularized stacked GRUs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_epochs <- 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model <- keras_model_sequential() %>%\n",
    "    layer_gru(\n",
    "        units = 32,\n",
    "        input_shape = list(NULL, dim(data)[[-1]]),\n",
    "        return_sequences = TRUE,\n",
    "        dropout = 0.1,\n",
    "        recurrent_dropout = 0.33\n",
    "    ) %>%\n",
    "    layer_gru(\n",
    "        units = 32,\n",
    "        dropout = 0.1,\n",
    "        recurrent_dropout = 0.33\n",
    "    ) %>%\n",
    "    layer_dense(units = 1)\n",
    "cat(summary(model))\n",
    "\n",
    "progress <- Progress$new(steps_per_epoch, num_epochs, 10)\n",
    "\n",
    "model %>% compile(\n",
    "    optimizer = optimizer_rmsprop(),\n",
    "    loss = \"mae\"\n",
    ")\n",
    "\n",
    "history <- model %>% fit_generator(\n",
    "    training_generator,\n",
    "    steps_per_epoch = steps_per_epoch,\n",
    "    epochs = num_epochs,\n",
    "    validation_data = validation_generator,\n",
    "    validation_steps = validation_steps,\n",
    "    callbacks = list(progress)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## DEBUG - stock prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<ol class=list-inline>\n",
       "\t<li>1239</li>\n",
       "\t<li>5</li>\n",
       "</ol>\n"
      ],
      "text/latex": [
       "\\begin{enumerate*}\n",
       "\\item 1239\n",
       "\\item 5\n",
       "\\end{enumerate*}\n"
      ],
      "text/markdown": [
       "1. 1239\n",
       "2. 5\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "[1] 1239    5"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th></th><th scope=col>close</th><th scope=col>volume</th><th scope=col>open</th><th scope=col>high</th><th scope=col>low</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>1239</th><td>558.46 </td><td>  13052</td><td>568.00 </td><td>568.00 </td><td>552.92 </td></tr>\n",
       "\t<tr><th scope=row>1238</th><td>559.99 </td><td>  41003</td><td>561.20 </td><td>566.43 </td><td>558.67 </td></tr>\n",
       "\t<tr><th scope=row>1237</th><td>556.97 </td><td>  10772</td><td>566.89 </td><td>567.00 </td><td>556.93 </td></tr>\n",
       "\t<tr><th scope=row>1236</th><td>567.16 </td><td>   7932</td><td>558.71 </td><td>568.45 </td><td>558.71 </td></tr>\n",
       "\t<tr><th scope=row>1235</th><td>567.00 </td><td> 146697</td><td>599.99 </td><td>604.83 </td><td>562.19 </td></tr>\n",
       "\t<tr><th scope=row>1234</th><td>569.74 </td><td>5080231</td><td>569.85 </td><td>587.28 </td><td>564.13 </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllll}\n",
       "  & close & volume & open & high & low\\\\\n",
       "\\hline\n",
       "\t1239 & 558.46  &   13052 & 568.00  & 568.00  & 552.92 \\\\\n",
       "\t1238 & 559.99  &   41003 & 561.20  & 566.43  & 558.67 \\\\\n",
       "\t1237 & 556.97  &   10772 & 566.89  & 567.00  & 556.93 \\\\\n",
       "\t1236 & 567.16  &    7932 & 558.71  & 568.45  & 558.71 \\\\\n",
       "\t1235 & 567.00  &  146697 & 599.99  & 604.83  & 562.19 \\\\\n",
       "\t1234 & 569.74  & 5080231 & 569.85  & 587.28  & 564.13 \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| <!--/--> | close | volume | open | high | low | \n",
       "|---|---|---|---|---|---|\n",
       "| 1239 | 558.46  |   13052 | 568.00  | 568.00  | 552.92  | \n",
       "| 1238 | 559.99  |   41003 | 561.20  | 566.43  | 558.67  | \n",
       "| 1237 | 556.97  |   10772 | 566.89  | 567.00  | 556.93  | \n",
       "| 1236 | 567.16  |    7932 | 558.71  | 568.45  | 558.71  | \n",
       "| 1235 | 567.00  |  146697 | 599.99  | 604.83  | 562.19  | \n",
       "| 1234 | 569.74  | 5080231 | 569.85  | 587.28  | 564.13  | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "     close  volume  open   high   low   \n",
       "1239 558.46   13052 568.00 568.00 552.92\n",
       "1238 559.99   41003 561.20 566.43 558.67\n",
       "1237 556.97   10772 566.89 567.00 556.93\n",
       "1236 567.16    7932 558.71 568.45 558.71\n",
       "1235 567.00  146697 599.99 604.83 562.19\n",
       "1234 569.74 5080231 569.85 587.28 564.13"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df <- read.csv(\"data/AAPL.csv\")\n",
    "df <- df[nrow(df):1,-1]\n",
    "mins <- apply(df, 2, min)\n",
    "maxs <- apply(df, 2, max)\n",
    "dim(df)\n",
    "head(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th></th><th scope=col>close</th><th scope=col>volume</th><th scope=col>open</th><th scope=col>high</th><th scope=col>low</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>1239</th><td>0.08495965  </td><td>0.0004593754</td><td>0.09448058  </td><td>0.09258607  </td><td>0.08583511  </td></tr>\n",
       "\t<tr><th scope=row>1238</th><td>0.08693186  </td><td>0.0029671884</td><td>0.08572165  </td><td>0.09056785  </td><td>0.09338639  </td></tr>\n",
       "\t<tr><th scope=row>1237</th><td>0.08303901  </td><td>0.0002548098</td><td>0.09305081  </td><td>0.09130058  </td><td>0.09110131  </td></tr>\n",
       "\t<tr><th scope=row>1236</th><td>0.09617417  </td><td>0.0000000000</td><td>0.08251433  </td><td>0.09316454  </td><td>0.09343892  </td></tr>\n",
       "\t<tr><th scope=row>1235</th><td>0.09596793  </td><td>0.0124502403</td><td>0.13568622  </td><td>0.13993064  </td><td>0.09800909  </td></tr>\n",
       "\t<tr><th scope=row>1234</th><td>0.09949986  </td><td>0.4550956039</td><td>0.09686353  </td><td>0.11737030  </td><td>0.10055683  </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllll}\n",
       "  & close & volume & open & high & low\\\\\n",
       "\\hline\n",
       "\t1239 & 0.08495965   & 0.0004593754 & 0.09448058   & 0.09258607   & 0.08583511  \\\\\n",
       "\t1238 & 0.08693186   & 0.0029671884 & 0.08572165   & 0.09056785   & 0.09338639  \\\\\n",
       "\t1237 & 0.08303901   & 0.0002548098 & 0.09305081   & 0.09130058   & 0.09110131  \\\\\n",
       "\t1236 & 0.09617417   & 0.0000000000 & 0.08251433   & 0.09316454   & 0.09343892  \\\\\n",
       "\t1235 & 0.09596793   & 0.0124502403 & 0.13568622   & 0.13993064   & 0.09800909  \\\\\n",
       "\t1234 & 0.09949986   & 0.4550956039 & 0.09686353   & 0.11737030   & 0.10055683  \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| <!--/--> | close | volume | open | high | low | \n",
       "|---|---|---|---|---|---|\n",
       "| 1239 | 0.08495965   | 0.0004593754 | 0.09448058   | 0.09258607   | 0.08583511   | \n",
       "| 1238 | 0.08693186   | 0.0029671884 | 0.08572165   | 0.09056785   | 0.09338639   | \n",
       "| 1237 | 0.08303901   | 0.0002548098 | 0.09305081   | 0.09130058   | 0.09110131   | \n",
       "| 1236 | 0.09617417   | 0.0000000000 | 0.08251433   | 0.09316454   | 0.09343892   | \n",
       "| 1235 | 0.09596793   | 0.0124502403 | 0.13568622   | 0.13993064   | 0.09800909   | \n",
       "| 1234 | 0.09949986   | 0.4550956039 | 0.09686353   | 0.11737030   | 0.10055683   | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "     close      volume       open       high       low       \n",
       "1239 0.08495965 0.0004593754 0.09448058 0.09258607 0.08583511\n",
       "1238 0.08693186 0.0029671884 0.08572165 0.09056785 0.09338639\n",
       "1237 0.08303901 0.0002548098 0.09305081 0.09130058 0.09110131\n",
       "1236 0.09617417 0.0000000000 0.08251433 0.09316454 0.09343892\n",
       "1235 0.09596793 0.0124502403 0.13568622 0.13993064 0.09800909\n",
       "1234 0.09949986 0.4550956039 0.09686353 0.11737030 0.10055683"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df <- t(apply(df, 1, function(x) (x - mins) / (maxs - mins)))\n",
    "head(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAHgCAMAAAC7G6qeAAADAFBMVEUAAAABAQECAgIDAwME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\nXJ8AAAAASUVORK5CYII=",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ggplot(data = as.data.frame(df), aes(x = 1:nrow(df), y = high)) + geom_line()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "generator2 <- function(data, lookback, delay, min_index, max_index, shuffle = FALSE, batch_size = 512, step = 6) {\n",
    "    if (is.null(max_index))\n",
    "        max_index <- nrow(data) - delay - 1\n",
    "    i <- min_index + lookback  # for looping through the data from start to end and back again\n",
    "    \n",
    "    function() {\n",
    "        if (shuffle)\n",
    "            rows <- sample(c((min_index + lookback):max_index), size = batch_size)\n",
    "        else {\n",
    "            # Loop around to start if necessary\n",
    "            if (i + batch_size >= max_index)\n",
    "                i <<- min_index + lookback\n",
    "            \n",
    "            # Select next batch_size rows\n",
    "            rows <- c(i:min(i + batch_size, max_index))\n",
    "            i <<- i + length(rows)\n",
    "        }\n",
    "        \n",
    "        samples <- array(0, dim = c(length(rows), lookback / step, dim(data)[[-1]]))\n",
    "        targets <- array(0, dim = c(length(rows)))\n",
    "        \n",
    "        for (j in 1:length(rows)) {\n",
    "            indices <- seq(rows[[j]] - lookback, rows[[j]] - 1, length.out = dim(samples)[[2]])\n",
    "            samples[j,,] <- data[indices,]\n",
    "            targets[[j]] <- data[rows[[j]] + delay, 1]\n",
    "        }\n",
    "        list(samples, targets)\n",
    "    }\n",
    "}\n",
    "\n",
    "lookback <- 14\n",
    "step <- 1\n",
    "delay <- 5\n",
    "batch_size <- 64\n",
    "steps_per_epoch <- 250\n",
    "num_epochs <- 25\n",
    "\n",
    "training_generator <- generator2(\n",
    "    df,\n",
    "    lookback = lookback,\n",
    "    delay = delay,\n",
    "    min_index = 1,\n",
    "    max_index = 1100,\n",
    "    shuffle = TRUE,\n",
    "    step = step,\n",
    "    batch_size = batch_size\n",
    ")\n",
    "\n",
    "validation_generator <- generator2(\n",
    "    df,\n",
    "    lookback = lookback,\n",
    "    delay = delay,\n",
    "    min_index = 1100,\n",
    "    max_index = 1238,\n",
    "    step = step,\n",
    "    batch_size = batch_size\n",
    ")\n",
    "\n",
    "# validation_steps <- (1239 - 1100 - lookback) / batch_size\n",
    "validation_steps <- 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "________________________________________________________________________________\n",
      "Layer (type)                        Output Shape                    Param #     \n",
      "================================================================================\n",
      "gru_13 (GRU)                        (None, None, 64)                13440       \n",
      "________________________________________________________________________________\n",
      "gru_14 (GRU)                        (None, 64)                      24768       \n",
      "________________________________________________________________________________\n",
      "dense_8 (Dense)                     (None, 1)                       65          \n",
      "================================================================================\n",
      "Total params: 38,273\n",
      "Trainable params: 38,273\n",
      "Non-trainable params: 0\n",
      "________________________________________________________________________________\n",
      "Epoch 1/25: 250/250 - loss: 0.0586181 - validation loss: 0.138209\n",
      "Epoch 2/25: 250/250 - loss: 0.03937405 - validation loss: 0.1119624\n",
      "Epoch 3/25: 250/250 - loss: 0.03708661 - validation loss: 0.09616219\n",
      "Epoch 4/25: 250/250 - loss: 0.03594268 - validation loss: 0.05729083\n",
      "Epoch 5/25: 250/250 - loss: 0.03493163 - validation loss: 0.1027096\n",
      "Epoch 6/25: 250/250 - loss: 0.03430449 - validation loss: 0.08497231\n",
      "Epoch 7/25: 250/250 - loss: 0.03456372 - validation loss: 0.1290234\n",
      "Epoch 8/25: 250/250 - loss: 0.03328745 - validation loss: 0.09239631\n",
      "Epoch 9/25: 250/250 - loss: 0.03295605 - validation loss: 0.09284522\n",
      "Epoch 10/25: 250/250 - loss: 0.03351588 - validation loss: 0.1270327\n",
      "Epoch 11/25: 250/250 - loss: 0.03335148 - validation loss: 0.08080583\n",
      "Epoch 12/25: 250/250 - loss: 0.03363662 - validation loss: 0.08111417\n",
      "Epoch 13/25: 250/250 - loss: 0.03284972 - validation loss: 0.1052079\n",
      "Epoch 14/25: 250/250 - loss: 0.03274519 - validation loss: 0.1122802\n",
      "Epoch 15/25: 250/250 - loss: 0.03389543 - validation loss: 0.07701647\n",
      "Epoch 16/25: 250/250 - loss: 0.04690994 - validation loss: 0.7961185\n",
      "Epoch 17/25: 250/250 - loss: 0.1486496 - validation loss: 0.1496269\n",
      "Epoch 18/25: 250/250 - loss: 0.09161487 - validation loss: 0.07207692\n",
      "Epoch 19/25: 250/250 - loss: 0.0831496 - validation loss: 0.07988522\n",
      "Epoch 20/25: 250/250 - loss: 0.07435886 - validation loss: 0.0544474\n",
      "Epoch 21/25: 250/250 - loss: 0.07356519 - validation loss: 0.05231667\n",
      "Epoch 22/25: 250/250 - loss: 0.07162824 - validation loss: 0.06004629\n",
      "Epoch 23/25: 250/250 - loss: 0.07113382 - validation loss: 0.05255947\n",
      "Epoch 24/25: 250/250 - loss: 0.06987641 - validation loss: 0.05550963\n",
      "Epoch 25/25: 250/250 - loss: 0.06702248 - validation loss: 0.05248111\n"
     ]
    }
   ],
   "source": [
    "model <- keras_model_sequential() %>%\n",
    "#     layer_simple_rnn(units = 32, input_shape = list(NULL, dim(data)[[-1]])) %>%\n",
    "#     layer_gru(units = 32, input_shape = list(NULL, dim(data)[[-1]])) %>%\n",
    "    layer_gru(units = 64, input_shape = list(NULL, dim(df)[[-1]]), recurrent_dropout = 0.2, dropout = 0.2, return_sequences = T) %>%\n",
    "    layer_gru(units = 64, recurrent_dropout = 0.2, dropout = 0.2) %>%\n",
    "#     layer_lstm(units = 32) %>%\n",
    "    layer_dense(units = 1)\n",
    "cat(summary(model))\n",
    "\n",
    "progress <- Progress$new(steps_per_epoch, num_epochs, 10)\n",
    "\n",
    "model %>% compile(\n",
    "    optimizer = optimizer_adam(lr = 0.01),\n",
    "    loss = \"mae\"\n",
    ")\n",
    "\n",
    "history <- model %>% fit_generator(\n",
    "    training_generator,\n",
    "    steps_per_epoch = steps_per_epoch,\n",
    "    epochs = num_epochs,\n",
    "    validation_data = validation_generator,\n",
    "    validation_steps = validation_steps,\n",
    "    callbacks = list(progress)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAHgCAIAAAADp837AAAACXBIWXMAABJ0AAASdAHeZh94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\nCCGEELK7/wfqyf7+paJR+gAAAABJRU5ErkJggg==",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## DEBUG - human activity classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>index</th><th scope=col>source</th><th scope=col>timestamp</th><th scope=col>values</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>1316828                             </td><td>step_detector                       </td><td>2017-06-29 07:45:16.506             </td><td>['1.0']                             </td></tr>\n",
       "\t<tr><td>1316829                             </td><td>battery                             </td><td>2017-06-29 07:45:16.514             </td><td>['100']                             </td></tr>\n",
       "\t<tr><td>1316830                             </td><td>activity                            </td><td>2017-06-29 07:45:16.579             </td><td>['STILL: 100']                      </td></tr>\n",
       "\t<tr><td>1316831                             </td><td>audio                               </td><td>2017-06-29 07:45:17.407             </td><td>['0.0', '0.0', '0.0', '32767.0']    </td></tr>\n",
       "\t<tr><td>1316832                             </td><td>audio                               </td><td>2017-06-29 07:45:18.407             </td><td>['227.0', '113.5', '227.0', '227.0']</td></tr>\n",
       "\t<tr><td>1316833                             </td><td>activity                            </td><td>2017-06-29 07:45:18.767             </td><td>['STILL: 100']                      </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llll}\n",
       " index & source & timestamp & values\\\\\n",
       "\\hline\n",
       "\t 1316828                                  & step\\_detector                          & 2017-06-29 07:45:16.506                  & {[}'1.0'{]}                             \\\\\n",
       "\t 1316829                                  & battery                                  & 2017-06-29 07:45:16.514                  & {[}'100'{]}                             \\\\\n",
       "\t 1316830                                  & activity                                 & 2017-06-29 07:45:16.579                  & {[}'STILL: 100'{]}                      \\\\\n",
       "\t 1316831                                  & audio                                    & 2017-06-29 07:45:17.407                  & {[}'0.0', '0.0', '0.0', '32767.0'{]}    \\\\\n",
       "\t 1316832                                  & audio                                    & 2017-06-29 07:45:18.407                  & {[}'227.0', '113.5', '227.0', '227.0'{]}\\\\\n",
       "\t 1316833                                  & activity                                 & 2017-06-29 07:45:18.767                  & {[}'STILL: 100'{]}                      \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "index | source | timestamp | values | \n",
       "|---|---|---|---|---|---|\n",
       "| 1316828                              | step_detector                        | 2017-06-29 07:45:16.506              | ['1.0']                              | \n",
       "| 1316829                              | battery                              | 2017-06-29 07:45:16.514              | ['100']                              | \n",
       "| 1316830                              | activity                             | 2017-06-29 07:45:16.579              | ['STILL: 100']                       | \n",
       "| 1316831                              | audio                                | 2017-06-29 07:45:17.407              | ['0.0', '0.0', '0.0', '32767.0']     | \n",
       "| 1316832                              | audio                                | 2017-06-29 07:45:18.407              | ['227.0', '113.5', '227.0', '227.0'] | \n",
       "| 1316833                              | activity                             | 2017-06-29 07:45:18.767              | ['STILL: 100']                       | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  index   source        timestamp              \n",
       "1 1316828 step_detector 2017-06-29 07:45:16.506\n",
       "2 1316829 battery       2017-06-29 07:45:16.514\n",
       "3 1316830 activity      2017-06-29 07:45:16.579\n",
       "4 1316831 audio         2017-06-29 07:45:17.407\n",
       "5 1316832 audio         2017-06-29 07:45:18.407\n",
       "6 1316833 activity      2017-06-29 07:45:18.767\n",
       "  values                              \n",
       "1 ['1.0']                             \n",
       "2 ['100']                             \n",
       "3 ['STILL: 100']                      \n",
       "4 ['0.0', '0.0', '0.0', '32767.0']    \n",
       "5 ['227.0', '113.5', '227.0', '227.0']\n",
       "6 ['STILL: 100']                      "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df <- read.csv('data/smartphone.csv', stringsAsFactors = F)\n",
    "head(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th></th><th scope=col>index</th><th scope=col>source</th><th scope=col>timestamp</th><th scope=col>values</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>18</th><td>1316845                                     </td><td>linear_acceleration                         </td><td>2017-06-29 07:45:22.007                     </td><td>['0.12779976', '-0.06277906', '0.45205125'] </td></tr>\n",
       "\t<tr><th scope=row>33</th><td>1316860                                     </td><td>linear_acceleration                         </td><td>2017-06-29 07:45:27.030                     </td><td>['-0.06043175', '0.029718934', '0.41986313']</td></tr>\n",
       "\t<tr><th scope=row>51</th><td>1316878                                     </td><td>linear_acceleration                         </td><td>2017-06-29 07:45:32.054                     </td><td>['0.84089065', '0.3555275', '0.52006143']   </td></tr>\n",
       "\t<tr><th scope=row>66</th><td>1316894                                     </td><td>linear_acceleration                         </td><td>2017-06-29 07:45:37.093                     </td><td>['0.5095852', '0.37330416', '0.23654178']   </td></tr>\n",
       "\t<tr><th scope=row>80</th><td>1316908                                     </td><td>linear_acceleration                         </td><td>2017-06-29 07:45:42.118                     </td><td>['0.46582192', '0.34427842', '0.0957648']   </td></tr>\n",
       "\t<tr><th scope=row>94</th><td>1316922                                     </td><td>linear_acceleration                         </td><td>2017-06-29 07:45:47.134                     </td><td>['0.5302881', '0.35416028', '0.21512994']   </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llll}\n",
       "  & index & source & timestamp & values\\\\\n",
       "\\hline\n",
       "\t18 & 1316845                                          & linear\\_acceleration                            & 2017-06-29 07:45:22.007                          & {[}'0.12779976', '-0.06277906', '0.45205125'{]} \\\\\n",
       "\t33 & 1316860                                          & linear\\_acceleration                            & 2017-06-29 07:45:27.030                          & {[}'-0.06043175', '0.029718934', '0.41986313'{]}\\\\\n",
       "\t51 & 1316878                                          & linear\\_acceleration                            & 2017-06-29 07:45:32.054                          & {[}'0.84089065', '0.3555275', '0.52006143'{]}   \\\\\n",
       "\t66 & 1316894                                          & linear\\_acceleration                            & 2017-06-29 07:45:37.093                          & {[}'0.5095852', '0.37330416', '0.23654178'{]}   \\\\\n",
       "\t80 & 1316908                                          & linear\\_acceleration                            & 2017-06-29 07:45:42.118                          & {[}'0.46582192', '0.34427842', '0.0957648'{]}   \\\\\n",
       "\t94 & 1316922                                          & linear\\_acceleration                            & 2017-06-29 07:45:47.134                          & {[}'0.5302881', '0.35416028', '0.21512994'{]}   \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| <!--/--> | index | source | timestamp | values | \n",
       "|---|---|---|---|---|---|\n",
       "| 18 | 1316845                                      | linear_acceleration                          | 2017-06-29 07:45:22.007                      | ['0.12779976', '-0.06277906', '0.45205125']  | \n",
       "| 33 | 1316860                                      | linear_acceleration                          | 2017-06-29 07:45:27.030                      | ['-0.06043175', '0.029718934', '0.41986313'] | \n",
       "| 51 | 1316878                                      | linear_acceleration                          | 2017-06-29 07:45:32.054                      | ['0.84089065', '0.3555275', '0.52006143']    | \n",
       "| 66 | 1316894                                      | linear_acceleration                          | 2017-06-29 07:45:37.093                      | ['0.5095852', '0.37330416', '0.23654178']    | \n",
       "| 80 | 1316908                                      | linear_acceleration                          | 2017-06-29 07:45:42.118                      | ['0.46582192', '0.34427842', '0.0957648']    | \n",
       "| 94 | 1316922                                      | linear_acceleration                          | 2017-06-29 07:45:47.134                      | ['0.5302881', '0.35416028', '0.21512994']    | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "   index   source              timestamp              \n",
       "18 1316845 linear_acceleration 2017-06-29 07:45:22.007\n",
       "33 1316860 linear_acceleration 2017-06-29 07:45:27.030\n",
       "51 1316878 linear_acceleration 2017-06-29 07:45:32.054\n",
       "66 1316894 linear_acceleration 2017-06-29 07:45:37.093\n",
       "80 1316908 linear_acceleration 2017-06-29 07:45:42.118\n",
       "94 1316922 linear_acceleration 2017-06-29 07:45:47.134\n",
       "   values                                      \n",
       "18 ['0.12779976', '-0.06277906', '0.45205125'] \n",
       "33 ['-0.06043175', '0.029718934', '0.41986313']\n",
       "51 ['0.84089065', '0.3555275', '0.52006143']   \n",
       "66 ['0.5095852', '0.37330416', '0.23654178']   \n",
       "80 ['0.46582192', '0.34427842', '0.0957648']   \n",
       "94 ['0.5302881', '0.35416028', '0.21512994']   "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df <- df[df['source'] == 'linear_acceleration',]\n",
    "head(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>x</th><th scope=col>y</th><th scope=col>z</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td> 0.12779976</td><td>-0.06277906</td><td>0.4520513  </td></tr>\n",
       "\t<tr><td>-0.06043175</td><td> 0.02971893</td><td>0.4198631  </td></tr>\n",
       "\t<tr><td> 0.84089065</td><td> 0.35552750</td><td>0.5200614  </td></tr>\n",
       "\t<tr><td> 0.50958520</td><td> 0.37330416</td><td>0.2365418  </td></tr>\n",
       "\t<tr><td> 0.46582192</td><td> 0.34427842</td><td>0.0957648  </td></tr>\n",
       "\t<tr><td> 0.53028810</td><td> 0.35416028</td><td>0.2151299  </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{lll}\n",
       " x & y & z\\\\\n",
       "\\hline\n",
       "\t  0.12779976 & -0.06277906 & 0.4520513  \\\\\n",
       "\t -0.06043175 &  0.02971893 & 0.4198631  \\\\\n",
       "\t  0.84089065 &  0.35552750 & 0.5200614  \\\\\n",
       "\t  0.50958520 &  0.37330416 & 0.2365418  \\\\\n",
       "\t  0.46582192 &  0.34427842 & 0.0957648  \\\\\n",
       "\t  0.53028810 &  0.35416028 & 0.2151299  \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "x | y | z | \n",
       "|---|---|---|---|---|---|\n",
       "|  0.12779976 | -0.06277906 | 0.4520513   | \n",
       "| -0.06043175 |  0.02971893 | 0.4198631   | \n",
       "|  0.84089065 |  0.35552750 | 0.5200614   | \n",
       "|  0.50958520 |  0.37330416 | 0.2365418   | \n",
       "|  0.46582192 |  0.34427842 | 0.0957648   | \n",
       "|  0.53028810 |  0.35416028 | 0.2151299   | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "     x           y           z        \n",
       "[1,]  0.12779976 -0.06277906 0.4520513\n",
       "[2,] -0.06043175  0.02971893 0.4198631\n",
       "[3,]  0.84089065  0.35552750 0.5200614\n",
       "[4,]  0.50958520  0.37330416 0.2365418\n",
       "[5,]  0.46582192  0.34427842 0.0957648\n",
       "[6,]  0.53028810  0.35416028 0.2151299"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "parseXYZ <- function(s) {\n",
    "    xyz <- as.numeric(strsplit(gsub(\"[' ]\", '', substr(s, 2, nchar(s) - 1)), ',')[[1]])\n",
    "    names(xyz) <- c('x', 'y', 'z')\n",
    "    xyz\n",
    "}\n",
    "xyz <- t(sapply(df$values, parseXYZ))\n",
    "rownames(xyz) <- c()\n",
    "head(xyz)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<ol class=list-inline>\n",
       "\t<li>99750</li>\n",
       "\t<li>3</li>\n",
       "</ol>\n"
      ],
      "text/latex": [
       "\\begin{enumerate*}\n",
       "\\item 99750\n",
       "\\item 3\n",
       "\\end{enumerate*}\n"
      ],
      "text/markdown": [
       "1. 99750\n",
       "2. 3\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "[1] 99750     3"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>x</th><th scope=col>y</th><th scope=col>z</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td> 0.12779976</td><td>-0.06277906</td><td>0.4520513  </td></tr>\n",
       "\t<tr><td>-0.06043175</td><td> 0.02971893</td><td>0.4198631  </td></tr>\n",
       "\t<tr><td> 0.84089065</td><td> 0.35552750</td><td>0.5200614  </td></tr>\n",
       "\t<tr><td> 0.50958520</td><td> 0.37330416</td><td>0.2365418  </td></tr>\n",
       "\t<tr><td> 0.46582192</td><td> 0.34427842</td><td>0.0957648  </td></tr>\n",
       "\t<tr><td> 0.53028810</td><td> 0.35416028</td><td>0.2151299  </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lll}\n",
       " x & y & z\\\\\n",
       "\\hline\n",
       "\t  0.12779976 & -0.06277906 & 0.4520513  \\\\\n",
       "\t -0.06043175 &  0.02971893 & 0.4198631  \\\\\n",
       "\t  0.84089065 &  0.35552750 & 0.5200614  \\\\\n",
       "\t  0.50958520 &  0.37330416 & 0.2365418  \\\\\n",
       "\t  0.46582192 &  0.34427842 & 0.0957648  \\\\\n",
       "\t  0.53028810 &  0.35416028 & 0.2151299  \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "x | y | z | \n",
       "|---|---|---|---|---|---|\n",
       "|  0.12779976 | -0.06277906 | 0.4520513   | \n",
       "| -0.06043175 |  0.02971893 | 0.4198631   | \n",
       "|  0.84089065 |  0.35552750 | 0.5200614   | \n",
       "|  0.50958520 |  0.37330416 | 0.2365418   | \n",
       "|  0.46582192 |  0.34427842 | 0.0957648   | \n",
       "|  0.53028810 |  0.35416028 | 0.2151299   | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  x           y           z        \n",
       "1  0.12779976 -0.06277906 0.4520513\n",
       "2 -0.06043175  0.02971893 0.4198631\n",
       "3  0.84089065  0.35552750 0.5200614\n",
       "4  0.50958520  0.37330416 0.2365418\n",
       "5  0.46582192  0.34427842 0.0957648\n",
       "6  0.53028810  0.35416028 0.2151299"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df <- as.data.frame(xyz)\n",
    "dim(df)\n",
    "head(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAHgCAIAAAADp837AAAACXBIWXMAABJ0AAASdAHeZh94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\nAIBzKDgAAACAc/8HE2I0O5YWHoQAAAAASUVORK5CYII=",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ggplot() +\n",
    "    geom_line(data = df, aes(x = 1:nrow(df), y = x, colour = \"red\")) +\n",
    "    geom_line(data = df, aes(x = 1:nrow(df), y = y, colour = \"blue\")) +\n",
    "    geom_line(data = df, aes(x = 1:nrow(df), y = z, colour = \"green\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAHgCAMAAAC7G6qeAAADAFBMVEUAAAABAQECAgIDAwME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",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ggplot() + geom_line(data = df[1:100,], aes(x = 1:100, y = z))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DEBUG: preprocessed data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>timestamp</th><th scope=col>x</th><th scope=col>y</th><th scope=col>z</th><th scope=col>label</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>2017-06-29 07:45:21.520</td><td>-0.8126313             </td><td> 1.608939              </td><td> 2.14533385            </td><td>STILL                  </td></tr>\n",
       "\t<tr><td>2017-06-29 07:45:26.542</td><td>-0.5300865             </td><td> 1.745943              </td><td> 2.14800323            </td><td>STILL                  </td></tr>\n",
       "\t<tr><td>2017-06-29 07:45:31.548</td><td>-0.5180120             </td><td>-1.225030              </td><td> 0.61977278            </td><td>STILL                  </td></tr>\n",
       "\t<tr><td>2017-06-29 07:45:36.552</td><td>-0.6798111             </td><td>-1.132576              </td><td>-0.07293608            </td><td>STILL                  </td></tr>\n",
       "\t<tr><td>2017-06-29 07:45:41.626</td><td>-0.6112276             </td><td>-1.116289              </td><td>-0.10363413            </td><td>STILL                  </td></tr>\n",
       "\t<tr><td>2017-06-29 07:45:46.648</td><td>-0.6493832             </td><td>-1.127307              </td><td>-0.08806267            </td><td>STILL                  </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllll}\n",
       " timestamp & x & y & z & label\\\\\n",
       "\\hline\n",
       "\t 2017-06-29 07:45:21.520 & -0.8126313              &  1.608939               &  2.14533385             & STILL                  \\\\\n",
       "\t 2017-06-29 07:45:26.542 & -0.5300865              &  1.745943               &  2.14800323             & STILL                  \\\\\n",
       "\t 2017-06-29 07:45:31.548 & -0.5180120              & -1.225030               &  0.61977278             & STILL                  \\\\\n",
       "\t 2017-06-29 07:45:36.552 & -0.6798111              & -1.132576               & -0.07293608             & STILL                  \\\\\n",
       "\t 2017-06-29 07:45:41.626 & -0.6112276              & -1.116289               & -0.10363413             & STILL                  \\\\\n",
       "\t 2017-06-29 07:45:46.648 & -0.6493832              & -1.127307               & -0.08806267             & STILL                  \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "timestamp | x | y | z | label | \n",
       "|---|---|---|---|---|---|\n",
       "| 2017-06-29 07:45:21.520 | -0.8126313              |  1.608939               |  2.14533385             | STILL                   | \n",
       "| 2017-06-29 07:45:26.542 | -0.5300865              |  1.745943               |  2.14800323             | STILL                   | \n",
       "| 2017-06-29 07:45:31.548 | -0.5180120              | -1.225030               |  0.61977278             | STILL                   | \n",
       "| 2017-06-29 07:45:36.552 | -0.6798111              | -1.132576               | -0.07293608             | STILL                   | \n",
       "| 2017-06-29 07:45:41.626 | -0.6112276              | -1.116289               | -0.10363413             | STILL                   | \n",
       "| 2017-06-29 07:45:46.648 | -0.6493832              | -1.127307               | -0.08806267             | STILL                   | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  timestamp               x          y         z           label\n",
       "1 2017-06-29 07:45:21.520 -0.8126313  1.608939  2.14533385 STILL\n",
       "2 2017-06-29 07:45:26.542 -0.5300865  1.745943  2.14800323 STILL\n",
       "3 2017-06-29 07:45:31.548 -0.5180120 -1.225030  0.61977278 STILL\n",
       "4 2017-06-29 07:45:36.552 -0.6798111 -1.132576 -0.07293608 STILL\n",
       "5 2017-06-29 07:45:41.626 -0.6112276 -1.116289 -0.10363413 STILL\n",
       "6 2017-06-29 07:45:46.648 -0.6493832 -1.127307 -0.08806267 STILL"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data <- read.csv('data/smartphone-preprocessed.csv')\n",
    "head(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "IN_VEHICLE ON_BICYCLE    ON_FOOT      STILL    TILTING    UNKNOWN \n",
       "      4679       1098       9788      72003          3      12296 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "table(data$label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "y\n",
       "    0     1     2     3     4     5 \n",
       " 4679  1098  9788 72003     3 12296 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y <- as.numeric(data$label) - 1\n",
    "table(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>x</th><th scope=col>y</th><th scope=col>z</th><th scope=col>label</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>-0.8126313 </td><td> 1.608939  </td><td> 2.14533385</td><td>3          </td></tr>\n",
       "\t<tr><td>-0.5300865 </td><td> 1.745943  </td><td> 2.14800323</td><td>3          </td></tr>\n",
       "\t<tr><td>-0.5180120 </td><td>-1.225030  </td><td> 0.61977278</td><td>3          </td></tr>\n",
       "\t<tr><td>-0.6798111 </td><td>-1.132576  </td><td>-0.07293608</td><td>3          </td></tr>\n",
       "\t<tr><td>-0.6112276 </td><td>-1.116289  </td><td>-0.10363413</td><td>3          </td></tr>\n",
       "\t<tr><td>-0.6493832 </td><td>-1.127307  </td><td>-0.08806267</td><td>3          </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{llll}\n",
       " x & y & z & label\\\\\n",
       "\\hline\n",
       "\t -0.8126313  &  1.608939   &  2.14533385 & 3          \\\\\n",
       "\t -0.5300865  &  1.745943   &  2.14800323 & 3          \\\\\n",
       "\t -0.5180120  & -1.225030   &  0.61977278 & 3          \\\\\n",
       "\t -0.6798111  & -1.132576   & -0.07293608 & 3          \\\\\n",
       "\t -0.6112276  & -1.116289   & -0.10363413 & 3          \\\\\n",
       "\t -0.6493832  & -1.127307   & -0.08806267 & 3          \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "x | y | z | label | \n",
       "|---|---|---|---|---|---|\n",
       "| -0.8126313  |  1.608939   |  2.14533385 | 3           | \n",
       "| -0.5300865  |  1.745943   |  2.14800323 | 3           | \n",
       "| -0.5180120  | -1.225030   |  0.61977278 | 3           | \n",
       "| -0.6798111  | -1.132576   | -0.07293608 | 3           | \n",
       "| -0.6112276  | -1.116289   | -0.10363413 | 3           | \n",
       "| -0.6493832  | -1.127307   | -0.08806267 | 3           | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "     x          y         z           label\n",
       "[1,] -0.8126313  1.608939  2.14533385 3    \n",
       "[2,] -0.5300865  1.745943  2.14800323 3    \n",
       "[3,] -0.5180120 -1.225030  0.61977278 3    \n",
       "[4,] -0.6798111 -1.132576 -0.07293608 3    \n",
       "[5,] -0.6112276 -1.116289 -0.10363413 3    \n",
       "[6,] -0.6493832 -1.127307 -0.08806267 3    "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df <- data\n",
    "df$label <- y\n",
    "df <- df[,-1]\n",
    "df <- as.matrix(df)\n",
    "head(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "99867"
      ],
      "text/latex": [
       "99867"
      ],
      "text/markdown": [
       "99867"
      ],
      "text/plain": [
       "[1] 99867"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "nrow(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "155"
      ],
      "text/latex": [
       "155"
      ],
      "text/markdown": [
       "155"
      ],
      "text/plain": [
       "[1] 155"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "generator3 <- function(data, lookback, delay, min_index, max_index, shuffle = FALSE, batch_size = 512, step = 6) {\n",
    "    if (is.null(max_index))\n",
    "        max_index <- nrow(data) - delay - 1\n",
    "    i <- min_index + lookback  # for looping through the data from start to end and back again\n",
    "    \n",
    "    function() {\n",
    "        if (shuffle)\n",
    "            rows <- sample(c((min_index + lookback):max_index), size = batch_size)\n",
    "        else {\n",
    "            # Loop around to start if necessary\n",
    "            if (i + batch_size >= max_index)\n",
    "                i <<- min_index + lookback\n",
    "            \n",
    "            # Select next batch_size rows\n",
    "            rows <- c(i:min(i + batch_size, max_index))\n",
    "            i <<- i + length(rows)\n",
    "        }\n",
    "        \n",
    "        samples <- array(0, dim = c(length(rows), lookback / step, dim(data)[[-1]] - 1))\n",
    "        targets <- array(0, dim = c(length(rows)))\n",
    "        \n",
    "        for (j in 1:length(rows)) {\n",
    "            indices <- seq(rows[[j]] - lookback, rows[[j]] - 1, length.out = dim(samples)[[2]])\n",
    "            samples[j,,] <- data[indices,-4]\n",
    "            targets[[j]] <- data[rows[[j]] + delay, 4]\n",
    "        }\n",
    "        \n",
    "        targets <- to_categorical(targets, num_classes = length(table(data[, 4])))\n",
    "        list(samples, targets)\n",
    "    }\n",
    "}\n",
    "\n",
    "lookback <- 60\n",
    "step <- 1\n",
    "delay <- 0\n",
    "batch_size <- 64\n",
    "steps_per_epoch <- 250\n",
    "num_epochs <- 25\n",
    "\n",
    "training_generator <- generator3(\n",
    "    df,\n",
    "    lookback = lookback,\n",
    "    delay = delay,\n",
    "    min_index = 1,\n",
    "    max_index = 70000,\n",
    "    shuffle = TRUE,\n",
    "    step = step,\n",
    "    batch_size = batch_size\n",
    ")\n",
    "\n",
    "validation_generator <- generator3(\n",
    "    df,\n",
    "    lookback = lookback,\n",
    "    delay = delay,\n",
    "    min_index = 70001,\n",
    "    max_index = 80000,\n",
    "    step = step,\n",
    "    batch_size = batch_size\n",
    ")\n",
    "\n",
    "validation_steps <- as.integer((80000 - 70001 - lookback) / batch_size)\n",
    "validation_steps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "________________________________________________________________________________\n",
      "Layer (type)                        Output Shape                    Param #     \n",
      "================================================================================\n",
      "gru_38 (GRU)                        (None, None, 64)                13056       \n",
      "________________________________________________________________________________\n",
      "gru_39 (GRU)                        (None, 64)                      24768       \n",
      "________________________________________________________________________________\n",
      "dense_32 (Dense)                    (None, 6)                       390         \n",
      "================================================================================\n",
      "Total params: 38,214\n",
      "Trainable params: 38,214\n",
      "Non-trainable params: 0\n",
      "________________________________________________________________________________\n",
      "Epoch 1/25: 250/250 - loss: 0.6985698, acc.: 0.7674375 - val. loss: 0.7591218 - val. acc.: 0.6661042\n",
      "Epoch 2/25: 250/250 - loss: 0.5167598, acc.: 0.8178125 - val. loss: 0.6949224 - val. acc.: 0.6946898\n",
      "Epoch 3/25: 250/250 - loss: 0.4666107, acc.: 0.8264375 - val. loss: 0.759693 - val. acc.: 0.66134\n",
      "Epoch 4/25: 250/250 - loss: 0.4443122, acc.: 0.835125 - val. loss: 0.7560308 - val. acc.: 0.6938958\n",
      "Epoch 5/25: 250/250 - loss: 0.4293655, acc.: 0.8401875 - val. loss: 0.7054694 - val. acc.: 0.7239702\n",
      "Epoch 6/25: 100/250 - loss: 0.4035773, acc.: 0.828125\r"
     ]
    }
   ],
   "source": [
    "model <- keras_model_sequential() %>%\n",
    "    # Flavour 1: Simple RNN\n",
    "#     layer_simple_rnn(units = 64, input_shape = list(NULL, dim(df)[[-1]] - 1), recurrent_dropout = 0.2, dropout = 0.2, return_sequences = T) %>%\n",
    "#     layer_simple_rnn(units = 64, recurrent_dropout = 0.2, dropout = 0.2) %>%\n",
    "\n",
    "    # Flavour 2: GRU: ~0.75\n",
    "#     layer_gru(units = 64, input_shape = list(NULL, dim(df)[[-1]] - 1), recurrent_dropout = 0.2, dropout = 0.2, return_sequences = T) %>%\n",
    "#     layer_gru(units = 64, recurrent_dropout = 0.2, dropout = 0.2) %>%\n",
    "\n",
    "    layer_gru(units = 64, input_shape = list(NULL, 3), return_sequences = T) %>%\n",
    "    layer_gru(units = 64) %>%\n",
    "\n",
    "    # Flavour 3: GRU + CNN\n",
    "#     layer_conv_1d(filters = 32, kernel_size = 3, activation = 'relu', input_shape = list(NULL, dim(df)[[-1]] - 1)) %>%\n",
    "#     layer_max_pooling_1d(pool_size = 2) %>%\n",
    "#     layer_conv_1d(filters = 64, kernel_size = 3, activation = 'relu') %>%\n",
    "#     layer_max_pooling_1d(pool_size = 2) %>%\n",
    "#     layer_gru(units = 64, recurrent_dropout = 0.2, dropout = 0.2, return_sequences = T) %>%\n",
    "#     layer_gru(units = 64, recurrent_dropout = 0.2, dropout = 0.2) %>\n",
    "    \n",
    "    # Flavour 4: CNN\n",
    "    \n",
    "#     layer_conv_1d(filters = 32, kernel_size = 3, activation = 'relu', input_shape = list(60, 3)) %>%\n",
    "#     layer_max_pooling_1d(pool_size = 2) %>%\n",
    "#     layer_conv_1d(filters = 64, kernel_size = 3, activation = 'relu') %>%\n",
    "#     layer_max_pooling_1d(pool_size = 2) %>%\n",
    "#     layer_conv_1d(filters = 128, kernel_size = 3, activation = 'relu') %>%\n",
    "#     layer_max_pooling_1d(pool_size = 2) %>%\n",
    "#     layer_flatten() %>%\n",
    "#     layer_dense(units = 256) %>%\n",
    "#     layer_dense(units = length(table(y)), activation = 'softmax')\n",
    "\n",
    "    # Flavour 5: DNN: ~ 60\n",
    "    \n",
    "#     layer_flatten(input_shape = list(60, 3)) %>%\n",
    "#     layer_dense(units = 256) %>%\n",
    "#     layer_dense(units = 512) %>%\n",
    "\n",
    "    layer_dense(units = length(table(y)), activation = 'softmax')\n",
    "\n",
    "cat(summary(model))\n",
    "\n",
    "progress <- Progress$new(steps_per_epoch, num_epochs, 10)\n",
    "\n",
    "model %>% compile(\n",
    "    optimizer = optimizer_adam(),\n",
    "    loss = \"categorical_crossentropy\",\n",
    "    metrics = c(\"accuracy\")\n",
    ")\n",
    "\n",
    "history <- model %>% fit_generator(\n",
    "    training_generator,\n",
    "    steps_per_epoch = steps_per_epoch,\n",
    "    epochs = num_epochs,\n",
    "    validation_data = validation_generator,\n",
    "    validation_steps = validation_steps,\n",
    "    callbacks = list(progress)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAHgCAIAAAADp837AAAACXBIWXMAABJ0AAASdAHeZh94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\nIIQQQsjr/h+cT6DiQfCXAgAAAABJRU5ErkJggg==",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAHgCAIAAAADp837AAAACXBIWXMAABJ0AAASdAHeZh94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\nQ8IBAAAAzCHhAAAAAOaQcAAAAABz/x/9F/7+Q3tBnAAAAABJRU5ErkJggg==",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_timeseries <- function(df, boxes) {\n",
    "    p <- ggplot() +\n",
    "        geom_line(data = df, aes(x = 1:nrow(df), y = x, colour = \"red\")) +\n",
    "        geom_line(data = df, aes(x = 1:nrow(df), y = y, colour = \"blue\")) +\n",
    "        geom_line(data = df, aes(x = 1:nrow(df), y = z, colour = \"green\")) +\n",
    "        geom_rect(data = boxes, mapping = aes(xmin = x1, xmax = x2, ymin = y1, ymax = y2, fill = as.factor(label)))\n",
    "    p\n",
    "}\n",
    "plot_timeseries(as.data.frame(df), boxes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>x1</th><th scope=col>x2</th><th scope=col>y1</th><th scope=col>y2</th><th scope=col>label</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>  1</td><td>699</td><td>9  </td><td>10 </td><td>3  </td></tr>\n",
       "\t<tr><td>699</td><td>706</td><td>9  </td><td>10 </td><td>2  </td></tr>\n",
       "\t<tr><td>706</td><td>710</td><td>9  </td><td>10 </td><td>5  </td></tr>\n",
       "\t<tr><td>710</td><td>713</td><td>9  </td><td>10 </td><td>2  </td></tr>\n",
       "\t<tr><td>713</td><td>719</td><td>9  </td><td>10 </td><td>5  </td></tr>\n",
       "\t<tr><td>719</td><td>729</td><td>9  </td><td>10 </td><td>3  </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllll}\n",
       " x1 & x2 & y1 & y2 & label\\\\\n",
       "\\hline\n",
       "\t   1 & 699 & 9   & 10  & 3  \\\\\n",
       "\t 699 & 706 & 9   & 10  & 2  \\\\\n",
       "\t 706 & 710 & 9   & 10  & 5  \\\\\n",
       "\t 710 & 713 & 9   & 10  & 2  \\\\\n",
       "\t 713 & 719 & 9   & 10  & 5  \\\\\n",
       "\t 719 & 729 & 9   & 10  & 3  \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "x1 | x2 | y1 | y2 | label | \n",
       "|---|---|---|---|---|---|\n",
       "|   1 | 699 | 9   | 10  | 3   | \n",
       "| 699 | 706 | 9   | 10  | 2   | \n",
       "| 706 | 710 | 9   | 10  | 5   | \n",
       "| 710 | 713 | 9   | 10  | 2   | \n",
       "| 713 | 719 | 9   | 10  | 5   | \n",
       "| 719 | 729 | 9   | 10  | 3   | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  x1  x2  y1 y2 label\n",
       "1   1 699 9  10 3    \n",
       "2 699 706 9  10 2    \n",
       "3 706 710 9  10 5    \n",
       "4 710 713 9  10 2    \n",
       "5 713 719 9  10 5    \n",
       "6 719 729 9  10 3    "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "last_index <- 1\n",
    "last_label <- as.numeric(df[1,4])\n",
    "\n",
    "boxes <- data.frame(x1=c(), x2=c(), y1=c(), y2=c(), stringsAsFactors = F)\n",
    "\n",
    "for (i in 2:nrow(df)) {\n",
    "    label <- as.numeric(df[i, 4])\n",
    "    if (label != last_label) {\n",
    "#         cat(last_index, \":\", i, \"->\", last_label, \"\\n\")\n",
    "        boxes <- rbind(boxes, data.frame(\n",
    "            x1=c(last_index),\n",
    "            x2=c(i),\n",
    "            y1=c(9),\n",
    "            y2=c(10),\n",
    "            label=c(last_label)\n",
    "        ))\n",
    "        last_label <- label\n",
    "        last_index <- i\n",
    "    }\n",
    "}\n",
    "head(boxes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<ol class=list-inline>\n",
       "\t<li>64</li>\n",
       "\t<li>10</li>\n",
       "\t<li>3</li>\n",
       "</ol>\n"
      ],
      "text/latex": [
       "\\begin{enumerate*}\n",
       "\\item 64\n",
       "\\item 10\n",
       "\\item 3\n",
       "\\end{enumerate*}\n"
      ],
      "text/markdown": [
       "1. 64\n",
       "2. 10\n",
       "3. 3\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "[1] 64 10  3"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dim(training_generator()[[1]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "y\n",
       "           0            1            2            3            4            5 \n",
       " 4.685231358  1.099462285  9.801035377 72.098891526  0.003003995 12.312375459 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(table(y) / nrow(df)) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "           0            1            2            3            4            5 \n",
       " 3.900000000  1.504285714  9.205714286 75.575714286  0.002857143  9.811428571 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(table(y[1:70000]) / nrow(df[1:70000,])) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "        0         1         2         3         4         5 \n",
       " 5.759424  0.409959 14.048595 59.044096  0.009999 20.727927 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(table(y[70000:80000]) / nrow(df[70000:80000,])) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "          0           1           2           3           5 \n",
       " 6.91061003  0.02013288  9.75941212 66.42339440 16.88645057 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(table(y[80000:99867]) / nrow(df[80000:99867,])) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "R",
   "language": "R",
   "name": "ir"
  },
  "language_info": {
   "codemirror_mode": "r",
   "file_extension": ".r",
   "mimetype": "text/x-r-source",
   "name": "R",
   "pygments_lexer": "r",
   "version": "3.5.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}