{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from utils.ipynb_helpers import read_data\n",
    "import numpy as np\n",
    "from scipy import stats\n",
    "import statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = read_data(\"./data/stock/full_1min.csv\")\n",
    "\n",
    "ticker = \"WTI\"\n",
    "stat = \"logpctchange\"\n",
    "df_t = df[ticker, stat]\n",
    "df_t_log = np.log(df_t + 1)\n",
    "df_t_exp = np.exp(df_t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_t[df.index > '2019'].plot()\n",
    "df_t.plot()\n",
    "pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Figure out stats for each ticker\n",
    "for t in df.columns.levels[0].to_list():\n",
    "    looking_at = df[t, stat][df.index < \"2023-01-01\"]\n",
    "    mean = statistics.mean(looking_at)\n",
    "    sd = statistics.stdev(looking_at)\n",
    "    print(f\"{t}:\\t\", \"mean:\", f\"{mean:.8f}\", \"sd:\", f\"{np.abs(mean):.6f}\", \"var:\", f\"{np.square(sd):.8f}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "looking_at = df[ticker, stat][df.index > \"2020-01-01\"]\n",
    "print(\"variance\", looking_at.var())\n",
    "mean = statistics.mean(looking_at)\n",
    "sd = statistics.stdev(looking_at)\n",
    "print(\"mean:\", mean, \"sd:\", sd, \"var:\", np.square(sd))\n",
    "# 1.092799419696949e-05\n",
    "# looking_at.plot.hist(bins=500, figsize=(20, 10))\n",
    "((looking_at - mean) / sd).plot.hist(\n",
    "    bins=200,\n",
    "    figsize=(20, 10),\n",
    "    title=\"Histogram of XOM Log Percent Change after 2020-01-01\",\n",
    ")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "max(df_t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "looking_at = df_t[df.index < \"2020-01-01\"]\n",
    "# print(looking_at.var())\n",
    "\n",
    "mean = statistics.mean(looking_at)\n",
    "sd = statistics.stdev(looking_at)\n",
    "print(mean, sd, np.square(sd))\n",
    "\n",
    "looking_at.plot.hist(\n",
    "    bins=2000,\n",
    "    figsize=(20, 10),\n",
    "    title=\"Histogram of XOM Log Percent Change after 2020-01-01\",\n",
    ")\n",
    "# ((looking_at-mean)/sd).plot.hist(bins=2000, figsize=(20, 10))\n",
    "# plt.plot((looking_at-mean)/sd)#stats.norm.pdf(looking_at, mean, sd)\n",
    "\n",
    "# df_t_log.plot.hist(bins=500, figsize=(20, 10))\n",
    "# df_t_exp.plot.hist(bins=500, figsize=(20, 10))\n",
    "pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Outlier stuff\n",
    "\n",
    "df_t_out = df_t[(np.abs(stats.zscore(df_t)) < 3)]\n",
    "print(len(df_t_out))\n",
    "df_t_out.plot.hist(bins=2000, figsize=(20, 10))\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "lower = df.XOM.pctchange.quantile(0.005)\n",
    "upper = df.XOM.pctchange.quantile(0.995)\n",
    "print(lower, upper)\n",
    "# # df.XOM.pctchange[df.XOM.pctchange.between(lower,upper)]\n",
    "df.XOM.pctchange[df.XOM.pctchange.between(lower, upper) == False]\n",
    "# # df.XOM.pctchange.clip(lower=lower, upper=upper)\n",
    "\n",
    "lower = df.quantile(0.005)\n",
    "upper = df.quantile(0.995)\n",
    "df.clip(lower=lower, upper=upper, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "start_date = \"2022-01-03\"\n",
    "end_date = \"2022-01-04\"\n",
    "f1 = df_t[df.index > start_date]\n",
    "f2 = f1[f1.index < end_date]\n",
    "print(f2)\n",
    "# f = plt.figure()\n",
    "# f.set_figwidth(60)\n",
    "# f.set_figheight(20)\n",
    "plt.figure(figsize=(24,4))\n",
    "plt.plot(f2.index.to_numpy(), np.exp(np.cumsum(f2.to_numpy())))\n",
    "f2.head(40)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.10.6"
  },
  "vscode": {
   "interpreter": {
    "hash": "51980e48e269f7c05efac26b22569386591d7f1d45336266d53ed7fc3ab7efc6"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}