{"cells":[
 {
  "cell_type": "markdown",
  "metadata": {},
  "source": "### As several people have mentioned, it does seem as 'time' represent minutes. Let's see whether there are some trend or/and seasonality in 'accuracy'."
 },
 {
  "cell_type": "code",
  "execution_count": null,
  "metadata": {
   "collapsed": false
  },
  "outputs": [],
  "source": "%matplotlib inline\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport datetime\nimport numpy as np\n\ndef dateTimeMaker(times):\n    return datetime.datetime(year=2000, month=1, day=1) + datetime.timedelta(minutes=int(times))\n# 01-01-2000 is an arbitrary date"
 },
 {
  "cell_type": "code",
  "execution_count": null,
  "metadata": {
   "collapsed": false
  },
  "outputs": [],
  "source": "%%time\ndf= pd.read_csv('../input/train.csv')\ndf['datetime']=df['time'].apply(dateTimeMaker)\ndf['date']=df['datetime'].dt.date\ndf['minutes']=df['datetime'].dt.minute\ndf['hours']=df['datetime'].dt.hour\ndf['weekday']=df['datetime'].dt.weekday"
 },
 {
  "cell_type": "code",
  "execution_count": null,
  "metadata": {
   "collapsed": false
  },
  "outputs": [],
  "source": "f, axes = plt.subplots(2, 2, figsize=(16, 9), sharex=False)\nfor i, param in enumerate('date,minutes,hours,weekday'.split(',')):\n    row,col=divmod(i,2)[0],divmod(i,2)[1]\n    grouped = pd.groupby(df[['accuracy',param]],by=param)[['accuracy']].mean()\n    axes[row,col].plot(grouped.index, grouped['accuracy'])\n    axes[row,col].set_ylim([40, 100]) \n    axes[row,col].set_xlabel(param,size=15)\n    axes[row,0].set_ylabel(\"accuracy\",size=15)"
 },
 {
  "cell_type": "markdown",
  "metadata": {},
  "source": "### It looks like 'accuracy' have some trend and maybe seasonality. Not sure how to interpret it yet, but it's something to consider."
 }
],"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"}}, "nbformat": 4, "nbformat_minor": 0}