04-create-samples.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from joblib import Parallel, delayed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_pickle('./data/all_session_ids.pkl.gz')\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random samples"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Set 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_ids = df.sample(frac=0.01)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_ids.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_ids_dict = sample_ids.groupby('filepath')['session_id'].apply(lambda x: list(x)).to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_one_file(x, y):\n",
    "    return pd.read_csv(x)\\\n",
    "    .loc[lambda k: k.session_id.isin(y)]\\\n",
    "    .copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_data = Parallel(n_jobs=14)(delayed(load_one_file)(filepath, one_sample_set)\n",
    "                                 for filepath, one_sample_set in sample_ids_dict.items())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_data = pd.concat(sample_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_data.to_pickle('./data/training_set_sample_data_1.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Set 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_ids = df\\\n",
    "    .loc[lambda x: ~x.session_id.isin(sample_ids.session_id.tolist())]\\\n",
    "    .sample(n=1249507)\n",
    "sample_ids.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_ids_dict = sample_ids.groupby('filepath')['session_id'].apply(lambda x: list(x)).to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_data = Parallel(n_jobs=14)(delayed(load_one_file)(filepath, one_sample_set)\n",
    "                                 for filepath, one_sample_set in sample_ids_dict.items())\n",
    "sample_data = pd.concat(sample_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_data.to_pickle('./data/training_set_sample_data_2.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create samples of particular length"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note: Overall random sample set 2 is not being included in the following samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_ids = pd.read_pickle('./data/training_set_sample_ids_2.pkl.gz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for a_len in range(10, 21):\n",
    "    this_sample_ids = df\\\n",
    "        .loc[lambda x: ~x.session_id.isin(sample_ids.session_id.tolist())]\\\n",
    "        .loc[lambda x: x.session_length == str(a_len)]\\\n",
    "        .sample(n=2499014)\n",
    "    sample_ids_dict = this_sample_ids.groupby('filepath')['session_id'].apply(lambda x: list(x)).to_dict()\n",
    "    sample_data = Parallel(n_jobs=14)(delayed(load_one_file)(filepath, one_sample_set)\n",
    "                                      for filepath, one_sample_set in sample_ids_dict.items())\n",
    "    sample_data = pd.concat(sample_data)\n",
    "    sample_data.to_pickle('./data/final_samples/training_set_sample_data_l{}.pkl'.format(a_len))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}