{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Algorithm Tuning\n",
    "\n",
    "This notebook examines the algorithm tuning process."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import OrderedDict\n",
    "import pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "import skopt\n",
    "import matplotlib.pyplot as plt\n",
    "from skopt.plots import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_opt(ds, algo):\n",
    "    return skopt.load(f'data/{ds}/tuning/{algo}.opt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def opt_df(opt):\n",
    "    ys = pd.Series(opt.func_vals)\n",
    "    cols = OrderedDict()\n",
    "    for i, dim in enumerate(opt.space):\n",
    "        cols[f'p{i}'] = pd.Series([x[i] for x in opt.x_iters])\n",
    "    cols['y'] = -ys\n",
    "    if hasattr(opt, 'iter_time'):\n",
    "        cols['time'] = pd.Series(opt.iter_time).reindex(ys.index)\n",
    "    return pd.DataFrame(cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def improvement_threshold(opt_df, col='y'):\n",
    "    ymax = opt_df[col].max()\n",
    "    return ymax - ymax * 0.01"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## User-User CF\n",
    "\n",
    "How does the user-user CF tuning perform?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "az_uu_i = read_opt('AZ', 'user-user-imp')\n",
    "az_uu_e = read_opt('AZ', 'user-user')\n",
    "bx_uu_i = read_opt('BX-I', 'user-user')\n",
    "bx_uu_e = read_opt('BX-E', 'user-user')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "BX-E UU plots:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_convergence(bx_uu_e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_objective(bx_uu_e)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "BX-I UU:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_convergence(bx_uu_i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_objective(bx_uu_i)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ALS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "bx_als_e = read_opt('BX-E', 'als')\n",
    "bx_als_i = read_opt('BX-I', 'wrls')\n",
    "az_als_e = read_opt('AZ', 'als')\n",
    "az_als_i = read_opt('AZ', 'wrls-imp')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_convergence(bx_als_e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_convergence(bx_als_i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_objective(bx_als_i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_convergence(az_als_e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_objective(az_als_e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_convergence(az_als_i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_objective(az_als_i)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluation Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "bxe_df = opt_df(bx_als_e)\n",
    "bxi_df = opt_df(bx_als_i)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### BX-E"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "bxe_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.scatterplot(x='p0', y='time', data=bxe_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.axvline(improvement_threshold(bxe_df), 0, 1, color='grey')\n",
    "sns.scatterplot(x='y', y='time', data=bxe_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.scatterplot(x='y', y='time', data=bxe_df)\n",
    "plt.vlines(bxe_df.y.max() - 0.005, 0, bxe_df.time.max())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### BX-I"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "bxi_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.scatterplot(x='p0', y='time', data=bxi_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.axvline(improvement_threshold(bxi_df), 0, 1, color='black')\n",
    "sns.scatterplot(x='y', y='time', data=bxi_df)"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}