research_amp/cc/notebooks/Master_model_performance_analyser.old.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "4d75d00e",
   "metadata": {},
   "source": [
    "# Description"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f710a02",
   "metadata": {},
   "source": [
    "Compute and analyze model performance stats."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4549477a",
   "metadata": {},
   "source": [
    "# Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "3e783ed1",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:47.345010Z",
     "start_time": "2022-07-07T13:26:45.860615Z"
    }
   },
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "fbc596de",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:50.383585Z",
     "start_time": "2022-07-07T13:26:47.347439Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/app/dataflow/core/dag.py:15: TqdmExperimentalWarning: Using `tqdm.autonotebook.tqdm` in notebook mode. Use `tqdm.tqdm` instead to force console mode (e.g. in jupyter console)\n",
      "  from tqdm.autonotebook import tqdm\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[0m\u001b[36mINFO\u001b[0m: > cmd='/venv/lib/python3.8/site-packages/ipykernel_launcher.py -f /home/.local/share/jupyter/runtime/kernel-14b68de8-a164-4348-af3b-9d7cbabd422d.json'\n",
      "WARNING: Running in Jupyter\n"
     ]
    }
   ],
   "source": [
    "import datetime\n",
    "import logging\n",
    "from typing import Any, Callable, List\n",
    "\n",
    "import ipywidgets as widgets\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scipy.stats as st\n",
    "\n",
    "import core.config.config_ as cconconf\n",
    "import core.config.config_utils as ccocouti\n",
    "import core.statistics.requires_statsmodels as cstresta\n",
    "import core.statistics.sharpe_ratio as cstshrat\n",
    "import dataflow.model as dtfmod\n",
    "import helpers.hdbg as hdbg\n",
    "import helpers.henv as henv\n",
    "import helpers.hpandas as hpandas\n",
    "import helpers.hprint as hprint\n",
    "import im_v2.crypto_chassis.data.client as iccdc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "717d24a8",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:50.542579Z",
     "start_time": "2022-07-07T13:26:50.386482Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[0m\u001b[33mWARNING\u001b[0m: Logger already initialized: skipping\n",
      "INFO  # Git\n",
      "    branch_name='CMTask2294_Unify_ImClient_examples_for_CCXT_and_CryptoChassis'\n",
      "    hash='68b0d0925'\n",
      "    # Last commits:\n",
      "      *   68b0d0925 Nina Lee Merge branch 'master' into CMTask2294_Unify_ImClient_examples_for_CCXT_and_CryptoChassis (21 minutes ago) Thu Jul 7 13:05:40 2022  (HEAD -> CMTask2294_Unify_ImClient_examples_for_CCXT_and_CryptoChassis, origin/CMTask2294_Unify_ImClient_examples_for_CCXT_and_CryptoChassis)\n",
      "      |\\  \n",
      "      | * a61aa42eb DanilYachmenev Cmtask2266: add 1 month period for test (#2296)                   (31 minutes ago) Thu Jul 7 12:55:51 2022  (origin/master, origin/HEAD)\n",
      "      * | 099daa6ca Nina Lee updated calls                                                     (37 minutes ago) Thu Jul 7 12:49:29 2022           \n",
      "# Machine info\n",
      "    system=Linux\n",
      "    node name=a800a26c72d5\n",
      "    release=5.13.0-1022-aws\n",
      "    version=#24~20.04.1-Ubuntu SMP Thu Apr 7 22:10:15 UTC 2022\n",
      "    machine=x86_64\n",
      "    processor=x86_64\n",
      "    cpu count=8\n",
      "    cpu freq=scpufreq(current=2499.998, min=0.0, max=0.0)\n",
      "    memory=svmem(total=33294798848, available=20900671488, percent=37.2, used=11635621888, free=9298796544, active=6005682176, inactive=15637479424, buffers=934797312, cached=11425583104, shared=280768512, slab=1836097536)\n",
      "    disk usage=sdiskusage(total=156066836480, used=26743250944, free=129306808320, percent=17.1)\n",
      "# Packages\n",
      "    python: 3.8.10\n",
      "    gluonnlp: ?\n",
      "    gluonts: 0.6.7\n",
      "    joblib: 1.1.0\n",
      "    mxnet: 1.9.0\n",
      "    numpy: 1.22.3\n",
      "    pandas: 1.4.2\n",
      "    pyarrow: 7.0.0\n",
      "    scipy: 1.6.1\n",
      "    seaborn: 0.11.2\n",
      "    sklearn: 1.0.2\n",
      "    statsmodels: 0.13.2\n",
      "\u001b[33mWARNING\u001b[0m: Disabling annoying warnings\n"
     ]
    }
   ],
   "source": [
    "hdbg.init_logger(verbosity=logging.INFO)\n",
    "\n",
    "_LOG = logging.getLogger(__name__)\n",
    "\n",
    "_LOG.info(\"%s\", henv.get_system_signature()[0])\n",
    "\n",
    "hprint.config_notebook()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f7dd094b",
   "metadata": {},
   "source": [
    "# Configs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "be23c175",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:50.599235Z",
     "start_time": "2022-07-07T13:26:50.544930Z"
    }
   },
   "outputs": [],
   "source": [
    "def get_notebook_config() -> cconconf.Config:\n",
    "    \"\"\"\n",
    "    Get notebook specific config.\n",
    "    \"\"\"\n",
    "    config = cconconf.Config()\n",
    "    param_dict = {\n",
    "        \"im_client_params\": {\n",
    "            \"universe_version\": \"v2\",\n",
    "            \"resample_1min\": True,\n",
    "            \"dataset\": \"ohlcv\",\n",
    "            \"contract_type\": \"spot\",\n",
    "            \"data_snapshot\": \"20220530\",\n",
    "        },\n",
    "        \"data\": {\n",
    "            \"dir_name\": \"/shared_data/model/historical/experiment.E1a.crypto_chassis_v2-all.5T.2018_2022/tiled_results/\",\n",
    "            \"columns\": \"volume vwap vwap.ret_0 vwap.ret_0.vol_adj vwap.ret_0.vol_adj.c vwap.ret_0.vol_adj.shift_-2 vwap.ret_0.vol_adj.shift_-2_hat\".split(),\n",
    "            \"start_date\": datetime.date(2018, 1, 1),\n",
    "            \"end_date\": datetime.date(2022, 5, 1),\n",
    "        },\n",
    "        \"column_names\": {\n",
    "            \"asset_id\": \"asset_id\",\n",
    "            \"timestamp\": \"end_ts\",\n",
    "            \"volume\": \"volume\",\n",
    "            \"y\": \"vwap.ret_0.vol_adj.shift_-2\",\n",
    "            \"y_hat\": \"vwap.ret_0.vol_adj.shift_-2_hat\",\n",
    "            \"hit\": \"hit\",\n",
    "            \"trade_pnl\": \"trade_pnl\",\n",
    "        },\n",
    "        \"stats_kwargs\": {\n",
    "            \"quantile_ranks\": 10,\n",
    "            # 28500 is the number of 5-minute intervals in ATH in a year.\n",
    "            \"time_scaling\": 28500,\n",
    "            \"n_resamples\": 1000,\n",
    "            \"alpha\": 0.05,\n",
    "        },\n",
    "        \"plot_kwargs\": {\n",
    "            \"y_min_lim_hit_rate\": 49,\n",
    "            \"y_max_lim_hit_rate\": 54,\n",
    "            \"color\": \"C0\",\n",
    "            \"capsize\": 0.2,\n",
    "            \"xticks_rotation\": 70,\n",
    "        },\n",
    "    }\n",
    "    config = cconfig.Config.from_dict(param_dict)\n",
    "    return config"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "fde4341e",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:50.792360Z",
     "start_time": "2022-07-07T13:26:50.602164Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "im_client_params:\n",
      "  universe_version: v2\n",
      "  resample_1min: True\n",
      "  dataset: ohlcv\n",
      "  contract_type: spot\n",
      "  data_snapshot: 20220530\n",
      "data:\n",
      "  dir_name: /shared_data/model/historical/experiment.E1a.crypto_chassis_v2-all.5T.2018_2022/tiled_results/\n",
      "  columns: ['volume', 'vwap', 'vwap.ret_0', 'vwap.ret_0.vol_adj', 'vwap.ret_0.vol_adj.c', 'vwap.ret_0.vol_adj.shift_-2', 'vwap.ret_0.vol_adj.shift_-2_hat']\n",
      "  start_date: 2018-01-01\n",
      "  end_date: 2022-05-01\n",
      "column_names:\n",
      "  asset_id: asset_id\n",
      "  timestamp: end_ts\n",
      "  volume: volume\n",
      "  y: vwap.ret_0.vol_adj.shift_-2\n",
      "  y_hat: vwap.ret_0.vol_adj.shift_-2_hat\n",
      "  hit: hit\n",
      "  trade_pnl: trade_pnl\n",
      "stats_kwargs:\n",
      "  quantile_ranks: 10\n",
      "  time_scaling: 28500\n",
      "  n_resamples: 1000\n",
      "  alpha: 0.05\n",
      "plot_kwargs:\n",
      "  y_min_lim_hit_rate: 49\n",
      "  y_max_lim_hit_rate: 54\n",
      "  color: C0\n",
      "  capsize: 0.2\n",
      "  xticks_rotation: 70\n"
     ]
    }
   ],
   "source": [
    "config = get_notebook_config()\n",
    "print(config)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0b3b5de5",
   "metadata": {},
   "source": [
    "# Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "862ac8a8",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:50.859336Z",
     "start_time": "2022-07-07T13:26:50.795489Z"
    }
   },
   "outputs": [],
   "source": [
    "def load_predictions_df(config: cconconf.Config) -> pd.DataFrame:\n",
    "    \"\"\"\n",
    "    Get data with ML predictions.\n",
    "    \"\"\"\n",
    "    backtest_df_iter = dtfmod.yield_processed_parquet_tiles_by_year(\n",
    "        config[\"data\"][\"dir_name\"],\n",
    "        config[\"data\"][\"start_date\"],\n",
    "        config[\"data\"][\"end_date\"],\n",
    "        config[\"column_names\"][\"asset_id\"],\n",
    "        data_cols=config[\"data\"][\"columns\"],\n",
    "        asset_ids=None,\n",
    "    )\n",
    "    #\n",
    "    #\n",
    "    predict_df = pd.concat(backtest_df_iter)\n",
    "    predict_df = predict_df.sort_index()\n",
    "    return predict_df\n",
    "\n",
    "\n",
    "# TODO(Max): Move the code out of the lib so we can unit test,\n",
    "# e.g., we want to add (small) specific unit tests for hit.\n",
    "# TODO(Max): Harmonize the code with calculate_hit_rate and other code there.\n",
    "# E.g., factor out the piece of calculate_hit_rate that computes hit, etc.\n",
    "def preprocess_predictions_df(\n",
    "    config: cconconf.Config, predict_df: pd.DataFrame\n",
    ") -> pd.DataFrame:\n",
    "    \"\"\"\n",
    "    Preprocess data with ML predictions for analysis.\n",
    "\n",
    "    Input:\n",
    "\n",
    "    ```\n",
    "                        volume                  vwap\n",
    "    asset_id            1464553467  1467591036  1464553467  1467591036\n",
    "    end_ts\n",
    "    2018-01-01 09:35:00   314.0657     47.3976    729.7789  12887.3945\n",
    "    2018-01-01 09:40:00   178.6543     35.1098    731.0134  12913.6854\n",
    "    ```\n",
    "\n",
    "    Output:\n",
    "\n",
    "    ```\n",
    "                                              volume        vwap\n",
    "    end_ts                        asset_id\n",
    "    2018-01-01 09:35:00  binance::ETH_USDT  314.0657    729.7789\n",
    "                         binance::BTC_USDT   47.3976  12887.3945\n",
    "    2018-01-01 09:40:00  binance::ETH_USDT  178.6543    731.0134\n",
    "                         binance::BTC_USDT   35.1098  12913.6854\n",
    "    ```\n",
    "    \"\"\"\n",
    "    # Convert the prediction stats data to Multiindex by time and asset id.\n",
    "    metrics_df = predict_df.stack()\n",
    "    # Drop NaNs to compute the performance statistics.\n",
    "    metrics_df = hpandas.dropna(metrics_df, report_stats=True)\n",
    "    # Compute hit.\n",
    "    metrics_df[\"hit\"] = (\n",
    "        metrics_df[config[\"column_names\"][\"y\"]]\n",
    "        * metrics_df[config[\"column_names\"][\"y_hat\"]]\n",
    "        >= 0\n",
    "    )\n",
    "    # Convert hit rates to desired format (`calculate_hit_rate` input).\n",
    "    metrics_df[\"hit\"] = metrics_df[\"hit\"].replace(True, 1)\n",
    "    metrics_df[\"hit\"] = metrics_df[\"hit\"].replace(False, -1)\n",
    "    # Compute trade PnL.\n",
    "    metrics_df[\"trade_pnl\"] = (\n",
    "        metrics_df[config[\"column_names\"][\"y\"]]\n",
    "        * metrics_df[config[\"column_names\"][\"y_hat\"]]\n",
    "    )\n",
    "    # TODO(*): Think about avoiding using `ImClient` for mapping.\n",
    "    # Convert asset ids to full symbols using `ImClient` mapping.\n",
    "    im_client = iccdc.get_CryptoChassisHistoricalPqByTileClient_example1(\n",
    "        **config[\"im_client_params\"]\n",
    "    )\n",
    "    metrics_df.index = metrics_df.index.set_levels(\n",
    "        metrics_df.index.levels[1].map(\n",
    "            im_client._asset_id_to_full_symbol_mapping\n",
    "        ),\n",
    "        level=1,\n",
    "    )\n",
    "    return metrics_df\n",
    "\n",
    "\n",
    "# TODO(*): Consider using bootstraping function from SciPy\n",
    "#  https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.bootstrap.html.\n",
    "# TODO(Dan): Consider to return CI values.\n",
    "def bootstrap(\n",
    "    data: pd.Series, func: Callable, n_resamples: int = 100\n",
    ") -> List[Any]:\n",
    "    \"\"\"\n",
    "    Bootstrap computations on specified number of data resamples.\n",
    "\n",
    "    :param data: input data to resample\n",
    "    :param func: function accepting a series and returning a single scalar value\n",
    "    :param n_resamples: number of resamples to create\n",
    "    :return: bootstrapped computations\n",
    "    \"\"\"\n",
    "    res_list = []\n",
    "    for i in range(n_resamples):\n",
    "        resampled_data = data.sample(frac=1, replace=True)\n",
    "        res = func(resampled_data)\n",
    "        res_list.append(res)\n",
    "    return res_list\n",
    "\n",
    "\n",
    "def compute_sharpe_ratio(\n",
    "    config: cconconf.Config, metrics_df: pd.DataFrame, by_col: str\n",
    ") -> None:\n",
    "    \"\"\"\n",
    "    Compute Sharpe Ratio by specified column.\n",
    "    \"\"\"\n",
    "    res_list = []\n",
    "    for by, data in metrics_df.groupby(by_col):\n",
    "        srs = data[config[\"column_names\"][\"trade_pnl\"]].dropna()\n",
    "        func = lambda pnl: cstshrat.compute_sharpe_ratio(\n",
    "            pnl, time_scaling=config[\"stats_kwargs\"][\"time_scaling\"]\n",
    "        )\n",
    "        # Multiple Sharpe Ratios are being computed on many resamples\n",
    "        # in order to find and plot confidence intervals.\n",
    "        sharpe_ratio_srs = pd.Series(\n",
    "            bootstrap(srs, func, config[\"stats_kwargs\"][\"n_resamples\"]),\n",
    "            name=\"sharpe_ratio\",\n",
    "        )\n",
    "        # Transform and combine data for plotting.\n",
    "        sharpe_ratio_df = sharpe_ratio_srs.to_frame()\n",
    "        sharpe_ratio_df[by_col] = by\n",
    "        res_list.append(sharpe_ratio_df)\n",
    "    res_df = pd.concat(res_list)\n",
    "    #\n",
    "    return res_df\n",
    "\n",
    "\n",
    "def calculate_hit_rate_with_CI(\n",
    "    df: pd.DataFrame, group_by: str, value_col: str\n",
    ") -> pd.DataFrame:\n",
    "    \"\"\"\n",
    "    Compute hit rates, confidence intervals and errors relative to the specific\n",
    "    entity.\n",
    "\n",
    "    :param df: data with hit rates values\n",
    "    :param group_by: column name for grouping entity\n",
    "    :param value_col: column name for PnL data\n",
    "    :return: data with CIs and errors\n",
    "    \"\"\"\n",
    "    # Calculate mean value of statistics as well as CIs for each entity.\n",
    "    hit_df_stacked = df.groupby([group_by])[value_col].apply(\n",
    "        lambda data: cstresta.calculate_hit_rate(\n",
    "            data, alpha=config[\"stats_kwargs\"][\"alpha\"]\n",
    "        )\n",
    "    )\n",
    "    # Process the output and add errors.\n",
    "    hit_df = hit_df_stacked.unstack()\n",
    "    hit_errors_df = add_errors_to_ci_data(hit_df)\n",
    "    return hit_errors_df\n",
    "\n",
    "\n",
    "def calculate_CI_for_PnLs_or_SR(\n",
    "    df: pd.DataFrame, group_by: str, value_col: str\n",
    ") -> pd.DataFrame:\n",
    "    \"\"\"\n",
    "    Compute mean PnL or Sharpe Ratio, confidence intervals and errors relative\n",
    "    to the specific entity.\n",
    "\n",
    "    :param df: data with PnL or Sharpe Ratio values\n",
    "    :param group_by: column name for grouping entity\n",
    "    :param value_col: column name for PnL or Sharpe Ratio data\n",
    "    :return: data with CIs and errors\n",
    "    \"\"\"\n",
    "    grouper = df.groupby([group_by])[value_col]\n",
    "    # Calculate mean value of statistics for each entity.\n",
    "    pnl_df = grouper.mean().to_frame()\n",
    "    # Compute confidence intervals.\n",
    "    conf_ints = grouper.apply(\n",
    "        lambda data: st.t.interval(\n",
    "            1 - config[\"stats_kwargs\"][\"alpha\"],\n",
    "            data.size - 1,\n",
    "            np.mean(data),\n",
    "            st.sem(data),\n",
    "        )\n",
    "    )\n",
    "    # Attach confidence intervals to the mean value data.\n",
    "    pnl_df[[\"low\", \"high\"]] = pd.DataFrame(conf_ints.tolist(), index=pnl_df.index)\n",
    "    # Unify columns and calculate errors (required values for plotting).\n",
    "    pnl_errors_df = add_errors_to_ci_data(pnl_df)\n",
    "    return pnl_errors_df\n",
    "\n",
    "\n",
    "def add_errors_to_ci_data(df: pd.DataFrame) -> pd.DataFrame:\n",
    "    \"\"\"\n",
    "    Process CI data and add errors.\n",
    "    \"\"\"\n",
    "    # Unify columns for all plotting data.\n",
    "    df.columns = [\"y\", \"ci_low\", \"ci_high\"]\n",
    "    # Required values for plotting (`yerr` input).\n",
    "    df[\"errors\"] = (df[\"ci_high\"] - df[\"ci_low\"]) / 2\n",
    "    return df\n",
    "\n",
    "\n",
    "def plot_stats_barplot(\n",
    "    df: pd.DataFrame,\n",
    "    sort_by: str,\n",
    "    ascending: bool,\n",
    "    ylabel: str,\n",
    "    ylim_min: str,\n",
    "    ylim_max: str,\n",
    ") -> None:\n",
    "    \"\"\"\n",
    "    :param df: data with prediction statistics\n",
    "    :param sort_by: sorting parameter (e.g., by value, by asset, or None)\n",
    "    :param ylabel: name of the Y-axis graph\n",
    "    :param ylim_min: lower value on Y-axis graph scale\n",
    "    :param ylim_max: upper value on Y-axis graph scale\n",
    "    :return: barplot with model performance statistics\n",
    "    \"\"\"\n",
    "    # Sort data according to the input params.\n",
    "    if sort_by == \"x\":\n",
    "        df_sorted = df.sort_index(ascending=ascending)\n",
    "    elif not sort_by:\n",
    "        df_sorted = df.copy()\n",
    "    else:\n",
    "        df_sorted = df.sort_values(by=sort_by, ascending=ascending)\n",
    "    # Specify errors for plotting.\n",
    "    errors = df_sorted[\"errors\"]\n",
    "    # Plotting params.\n",
    "    df_sorted[\"y\"].plot.bar(\n",
    "        yerr=errors,\n",
    "        capsize=4,\n",
    "        width=0.8,\n",
    "    )\n",
    "    plt.xticks(rotation=xticks_rotation)\n",
    "    plt.ylabel(ylabel)\n",
    "    plt.ylim(ylim_min, ylim_max)\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "def plot_bars_with_widget(\n",
    "    df: pd.DataFrame,\n",
    "    ylabel: str,\n",
    "    ylim_min: float,\n",
    "    ylim_max: float,\n",
    ") -> None:\n",
    "    \"\"\"\n",
    "    Add widgets to expand the sorting parameters for barplots.\n",
    "\n",
    "    :param df: data with prediction statistics\n",
    "    :param ylabel: name of the Y-axis graph\n",
    "    :param ylim_min: lower value on Y-axis graph scale\n",
    "    :param ylim_max: upper value on Y-axis graph scale\n",
    "    :return: barplot with edible model performance statistics\n",
    "    \"\"\"\n",
    "    _ = widgets.interact(\n",
    "        plot_stats_barplot,\n",
    "        df=widgets.fixed(df),\n",
    "        sort_by=widgets.ToggleButtons(\n",
    "            options=[\"x\", \"y\", \"ci_low\", \"ci_high\", False], description=\"Sort by:\"\n",
    "        ),\n",
    "        ascending=widgets.ToggleButtons(\n",
    "            options=[True, False], description=\"Ascending:\"\n",
    "        ),\n",
    "        ylabel=widgets.fixed(ylabel),\n",
    "        ylim_min=widgets.FloatText(\n",
    "            value=ylim_min,\n",
    "            description=\"Min y-value:\",\n",
    "        ),\n",
    "        ylim_max=widgets.FloatText(\n",
    "            value=ylim_max,\n",
    "            description=\"Max y-value:\",\n",
    "        ),\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cd96a4ed",
   "metadata": {},
   "source": [
    "# Load data with predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "a776540a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:54.809174Z",
     "start_time": "2022-07-07T13:26:50.862022Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "be3feb5754204930b4ace834693fe132",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/5 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(123119, 224)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"3\" halign=\"left\">volume</th>\n",
       "      <th colspan=\"3\" halign=\"left\">vwap</th>\n",
       "      <th colspan=\"3\" halign=\"left\">vwap.ret_0</th>\n",
       "      <th colspan=\"3\" halign=\"left\">vwap.ret_0.vol_adj</th>\n",
       "      <th colspan=\"3\" halign=\"left\">vwap.ret_0.vol_adj.c</th>\n",
       "      <th colspan=\"3\" halign=\"left\">vwap.ret_0.vol_adj.shift_-2</th>\n",
       "      <th colspan=\"3\" halign=\"left\">vwap.ret_0.vol_adj.shift_-2_hat</th>\n",
       "      <th colspan=\"29\" halign=\"left\">volume</th>\n",
       "      <th colspan=\"29\" halign=\"left\">vwap</th>\n",
       "      <th colspan=\"29\" halign=\"left\">vwap.ret_0</th>\n",
       "      <th colspan=\"29\" halign=\"left\">vwap.ret_0.vol_adj</th>\n",
       "      <th colspan=\"29\" halign=\"left\">vwap.ret_0.vol_adj.c</th>\n",
       "      <th colspan=\"29\" halign=\"left\">vwap.ret_0.vol_adj.shift_-2</th>\n",
       "      <th colspan=\"29\" halign=\"left\">vwap.ret_0.vol_adj.shift_-2_hat</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>asset_id</th>\n",
       "      <th>1464553467</th>\n",
       "      <th>1467591036</th>\n",
       "      <th>2061507978</th>\n",
       "      <th>1464553467</th>\n",
       "      <th>1467591036</th>\n",
       "      <th>2061507978</th>\n",
       "      <th>1464553467</th>\n",
       "      <th>1467591036</th>\n",
       "      <th>2061507978</th>\n",
       "      <th>1464553467</th>\n",
       "      <th>1467591036</th>\n",
       "      <th>2061507978</th>\n",
       "      <th>1464553467</th>\n",
       "      <th>1467591036</th>\n",
       "      <th>2061507978</th>\n",
       "      <th>1464553467</th>\n",
       "      <th>1467591036</th>\n",
       "      <th>2061507978</th>\n",
       "      <th>1464553467</th>\n",
       "      <th>1467591036</th>\n",
       "      <th>2061507978</th>\n",
       "      <th>1032127330</th>\n",
       "      <th>1269521755</th>\n",
       "      <th>1345967712</th>\n",
       "      <th>1480530030</th>\n",
       "      <th>1579162287</th>\n",
       "      <th>1606590437</th>\n",
       "      <th>2002879833</th>\n",
       "      <th>2050321118</th>\n",
       "      <th>2061357471</th>\n",
       "      <th>2063960810</th>\n",
       "      <th>2152163375</th>\n",
       "      <th>2237530510</th>\n",
       "      <th>2425180446</th>\n",
       "      <th>2502756391</th>\n",
       "      <th>2568064341</th>\n",
       "      <th>2743007870</th>\n",
       "      <th>2870803583</th>\n",
       "      <th>2871263732</th>\n",
       "      <th>3065029174</th>\n",
       "      <th>3145747555</th>\n",
       "      <th>3187272957</th>\n",
       "      <th>3254199816</th>\n",
       "      <th>3303714233</th>\n",
       "      <th>3554906657</th>\n",
       "      <th>6132418113</th>\n",
       "      <th>6352701412</th>\n",
       "      <th>6514937930</th>\n",
       "      <th>7348328528</th>\n",
       "      <th>8968126878</th>\n",
       "      <th>1032127330</th>\n",
       "      <th>1269521755</th>\n",
       "      <th>1345967712</th>\n",
       "      <th>1480530030</th>\n",
       "      <th>1579162287</th>\n",
       "      <th>1606590437</th>\n",
       "      <th>2002879833</th>\n",
       "      <th>2050321118</th>\n",
       "      <th>2061357471</th>\n",
       "      <th>2063960810</th>\n",
       "      <th>2152163375</th>\n",
       "      <th>2237530510</th>\n",
       "      <th>2425180446</th>\n",
       "      <th>2502756391</th>\n",
       "      <th>2568064341</th>\n",
       "      <th>2743007870</th>\n",
       "      <th>2870803583</th>\n",
       "      <th>2871263732</th>\n",
       "      <th>3065029174</th>\n",
       "      <th>3145747555</th>\n",
       "      <th>3187272957</th>\n",
       "      <th>3254199816</th>\n",
       "      <th>3303714233</th>\n",
       "      <th>3554906657</th>\n",
       "      <th>6132418113</th>\n",
       "      <th>6352701412</th>\n",
       "      <th>6514937930</th>\n",
       "      <th>7348328528</th>\n",
       "      <th>8968126878</th>\n",
       "      <th>1032127330</th>\n",
       "      <th>1269521755</th>\n",
       "      <th>1345967712</th>\n",
       "      <th>1480530030</th>\n",
       "      <th>1579162287</th>\n",
       "      <th>1606590437</th>\n",
       "      <th>2002879833</th>\n",
       "      <th>2050321118</th>\n",
       "      <th>2061357471</th>\n",
       "      <th>2063960810</th>\n",
       "      <th>2152163375</th>\n",
       "      <th>2237530510</th>\n",
       "      <th>2425180446</th>\n",
       "      <th>2502756391</th>\n",
       "      <th>2568064341</th>\n",
       "      <th>2743007870</th>\n",
       "      <th>2870803583</th>\n",
       "      <th>2871263732</th>\n",
       "      <th>3065029174</th>\n",
       "      <th>3145747555</th>\n",
       "      <th>3187272957</th>\n",
       "      <th>3254199816</th>\n",
       "      <th>3303714233</th>\n",
       "      <th>3554906657</th>\n",
       "      <th>6132418113</th>\n",
       "      <th>6352701412</th>\n",
       "      <th>6514937930</th>\n",
       "      <th>7348328528</th>\n",
       "      <th>8968126878</th>\n",
       "      <th>1032127330</th>\n",
       "      <th>1269521755</th>\n",
       "      <th>1345967712</th>\n",
       "      <th>1480530030</th>\n",
       "      <th>1579162287</th>\n",
       "      <th>1606590437</th>\n",
       "      <th>2002879833</th>\n",
       "      <th>2050321118</th>\n",
       "      <th>2061357471</th>\n",
       "      <th>2063960810</th>\n",
       "      <th>2152163375</th>\n",
       "      <th>2237530510</th>\n",
       "      <th>2425180446</th>\n",
       "      <th>2502756391</th>\n",
       "      <th>2568064341</th>\n",
       "      <th>2743007870</th>\n",
       "      <th>2870803583</th>\n",
       "      <th>2871263732</th>\n",
       "      <th>3065029174</th>\n",
       "      <th>3145747555</th>\n",
       "      <th>3187272957</th>\n",
       "      <th>3254199816</th>\n",
       "      <th>3303714233</th>\n",
       "      <th>3554906657</th>\n",
       "      <th>6132418113</th>\n",
       "      <th>6352701412</th>\n",
       "      <th>6514937930</th>\n",
       "      <th>7348328528</th>\n",
       "      <th>8968126878</th>\n",
       "      <th>1032127330</th>\n",
       "      <th>1269521755</th>\n",
       "      <th>1345967712</th>\n",
       "      <th>1480530030</th>\n",
       "      <th>1579162287</th>\n",
       "      <th>1606590437</th>\n",
       "      <th>2002879833</th>\n",
       "      <th>2050321118</th>\n",
       "      <th>2061357471</th>\n",
       "      <th>2063960810</th>\n",
       "      <th>2152163375</th>\n",
       "      <th>2237530510</th>\n",
       "      <th>2425180446</th>\n",
       "      <th>2502756391</th>\n",
       "      <th>2568064341</th>\n",
       "      <th>2743007870</th>\n",
       "      <th>2870803583</th>\n",
       "      <th>2871263732</th>\n",
       "      <th>3065029174</th>\n",
       "      <th>3145747555</th>\n",
       "      <th>3187272957</th>\n",
       "      <th>3254199816</th>\n",
       "      <th>3303714233</th>\n",
       "      <th>3554906657</th>\n",
       "      <th>6132418113</th>\n",
       "      <th>6352701412</th>\n",
       "      <th>6514937930</th>\n",
       "      <th>7348328528</th>\n",
       "      <th>8968126878</th>\n",
       "      <th>1032127330</th>\n",
       "      <th>1269521755</th>\n",
       "      <th>1345967712</th>\n",
       "      <th>1480530030</th>\n",
       "      <th>1579162287</th>\n",
       "      <th>1606590437</th>\n",
       "      <th>2002879833</th>\n",
       "      <th>2050321118</th>\n",
       "      <th>2061357471</th>\n",
       "      <th>2063960810</th>\n",
       "      <th>2152163375</th>\n",
       "      <th>2237530510</th>\n",
       "      <th>2425180446</th>\n",
       "      <th>2502756391</th>\n",
       "      <th>2568064341</th>\n",
       "      <th>2743007870</th>\n",
       "      <th>2870803583</th>\n",
       "      <th>2871263732</th>\n",
       "      <th>3065029174</th>\n",
       "      <th>3145747555</th>\n",
       "      <th>3187272957</th>\n",
       "      <th>3254199816</th>\n",
       "      <th>3303714233</th>\n",
       "      <th>3554906657</th>\n",
       "      <th>6132418113</th>\n",
       "      <th>6352701412</th>\n",
       "      <th>6514937930</th>\n",
       "      <th>7348328528</th>\n",
       "      <th>8968126878</th>\n",
       "      <th>1032127330</th>\n",
       "      <th>1269521755</th>\n",
       "      <th>1345967712</th>\n",
       "      <th>1480530030</th>\n",
       "      <th>1579162287</th>\n",
       "      <th>1606590437</th>\n",
       "      <th>2002879833</th>\n",
       "      <th>2050321118</th>\n",
       "      <th>2061357471</th>\n",
       "      <th>2063960810</th>\n",
       "      <th>2152163375</th>\n",
       "      <th>2237530510</th>\n",
       "      <th>2425180446</th>\n",
       "      <th>2502756391</th>\n",
       "      <th>2568064341</th>\n",
       "      <th>2743007870</th>\n",
       "      <th>2870803583</th>\n",
       "      <th>2871263732</th>\n",
       "      <th>3065029174</th>\n",
       "      <th>3145747555</th>\n",
       "      <th>3187272957</th>\n",
       "      <th>3254199816</th>\n",
       "      <th>3303714233</th>\n",
       "      <th>3554906657</th>\n",
       "      <th>6132418113</th>\n",
       "      <th>6352701412</th>\n",
       "      <th>6514937930</th>\n",
       "      <th>7348328528</th>\n",
       "      <th>8968126878</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>end_ts</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-01-01 09:35:00-05:00</th>\n",
       "      <td>314.06346</td>\n",
       "      <td>47.446994</td>\n",
       "      <td>NaN</td>\n",
       "      <td>729.777396</td>\n",
       "      <td>12887.299954</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-01 09:40:00-05:00</th>\n",
       "      <td>178.63246</td>\n",
       "      <td>35.739673</td>\n",
       "      <td>NaN</td>\n",
       "      <td>731.019329</td>\n",
       "      <td>12913.685019</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.001700</td>\n",
       "      <td>0.002045</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-01 09:45:00-05:00</th>\n",
       "      <td>222.53335</td>\n",
       "      <td>42.027924</td>\n",
       "      <td>NaN</td>\n",
       "      <td>734.082241</td>\n",
       "      <td>12995.308947</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.004181</td>\n",
       "      <td>0.006301</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              volume                              vwap                          vwap.ret_0                       vwap.ret_0.vol_adj                       vwap.ret_0.vol_adj.c                       vwap.ret_0.vol_adj.shift_-2                       vwap.ret_0.vol_adj.shift_-2_hat                           volume                                                                                                                                                                                                                                                                                                                           vwap                                                                                                                                                                                                                                                                                                                     vwap.ret_0                                              \\\n",
       "asset_id                  1464553467 1467591036 2061507978  1464553467    1467591036 2061507978 1464553467 1467591036 2061507978         1464553467 1467591036 2061507978           1464553467 1467591036 2061507978           1464553467 1467591036 2061507978               1464553467 1467591036 2061507978 1032127330 1269521755 1345967712 1480530030 1579162287 1606590437 2002879833 2050321118 2061357471 2063960810 2152163375 2237530510 2425180446 2502756391 2568064341 2743007870 2870803583 2871263732 3065029174 3145747555 3187272957 3254199816 3303714233 3554906657 6132418113 6352701412 6514937930 7348328528 8968126878 1032127330 1269521755 1345967712 1480530030 1579162287 1606590437 2002879833 2050321118 2061357471 2063960810 2152163375 2237530510 2425180446 2502756391 2568064341 2743007870 2870803583 2871263732 3065029174 3145747555 3187272957 3254199816 3303714233 3554906657 6132418113 6352701412 6514937930 7348328528 8968126878 1032127330 1269521755 1345967712 1480530030 1579162287   \n",
       "end_ts                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                \n",
       "2018-01-01 09:35:00-05:00  314.06346  47.446994        NaN  729.777396  12887.299954        NaN        NaN        NaN        NaN                NaN        NaN        NaN                  NaN        NaN        NaN                  NaN        NaN        NaN                      NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN   \n",
       "2018-01-01 09:40:00-05:00  178.63246  35.739673        NaN  731.019329  12913.685019        NaN   0.001700   0.002045        NaN                NaN        NaN        NaN                  NaN        NaN        NaN                  NaN        NaN        NaN                      NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN   \n",
       "2018-01-01 09:45:00-05:00  222.53335  42.027924        NaN  734.082241  12995.308947        NaN   0.004181   0.006301        NaN                NaN        NaN        NaN                  NaN        NaN        NaN                  NaN        NaN        NaN                      NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN   \n",
       "\n",
       "                                                                                                                                                                                                                                                                                                  vwap.ret_0.vol_adj                                                                                                                                                                                                                                                                                                                     vwap.ret_0.vol_adj.c                                                                                                                                                                                                                                                                                                                     vwap.ret_0.vol_adj.shift_-2                        \\\n",
       "asset_id                  1606590437 2002879833 2050321118 2061357471 2063960810 2152163375 2237530510 2425180446 2502756391 2568064341 2743007870 2870803583 2871263732 3065029174 3145747555 3187272957 3254199816 3303714233 3554906657 6132418113 6352701412 6514937930 7348328528 8968126878         1032127330 1269521755 1345967712 1480530030 1579162287 1606590437 2002879833 2050321118 2061357471 2063960810 2152163375 2237530510 2425180446 2502756391 2568064341 2743007870 2870803583 2871263732 3065029174 3145747555 3187272957 3254199816 3303714233 3554906657 6132418113 6352701412 6514937930 7348328528 8968126878           1032127330 1269521755 1345967712 1480530030 1579162287 1606590437 2002879833 2050321118 2061357471 2063960810 2152163375 2237530510 2425180446 2502756391 2568064341 2743007870 2870803583 2871263732 3065029174 3145747555 3187272957 3254199816 3303714233 3554906657 6132418113 6352701412 6514937930 7348328528 8968126878           1032127330 1269521755 1345967712   \n",
       "end_ts                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         \n",
       "2018-01-01 09:35:00-05:00        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                  NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                  NaN        NaN        NaN   \n",
       "2018-01-01 09:40:00-05:00        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                  NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                  NaN        NaN        NaN   \n",
       "2018-01-01 09:45:00-05:00        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                  NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                  NaN        NaN        NaN   \n",
       "\n",
       "                                                                                                                                                                                                                                                                                                                        vwap.ret_0.vol_adj.shift_-2_hat                                                                                                                                                                                                                                                                                                                      \n",
       "asset_id                  1480530030 1579162287 1606590437 2002879833 2050321118 2061357471 2063960810 2152163375 2237530510 2425180446 2502756391 2568064341 2743007870 2870803583 2871263732 3065029174 3145747555 3187272957 3254199816 3303714233 3554906657 6132418113 6352701412 6514937930 7348328528 8968126878               1032127330 1269521755 1345967712 1480530030 1579162287 1606590437 2002879833 2050321118 2061357471 2063960810 2152163375 2237530510 2425180446 2502756391 2568064341 2743007870 2870803583 2871263732 3065029174 3145747555 3187272957 3254199816 3303714233 3554906657 6132418113 6352701412 6514937930 7348328528 8968126878  \n",
       "end_ts                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                \n",
       "2018-01-01 09:35:00-05:00        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                      NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN  \n",
       "2018-01-01 09:40:00-05:00        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                      NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN  \n",
       "2018-01-01 09:45:00-05:00        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN                      NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN        NaN  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predict_df = load_predictions_df(config)\n",
    "print(predict_df.shape)\n",
    "predict_df.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1f8d0c59",
   "metadata": {},
   "source": [
    "# Compute overall PnL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c0e507a4",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:56.341033Z",
     "start_time": "2022-07-07T13:26:54.811201Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='end_ts'>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(\n",
    "    predict_df[config[\"column_names\"][\"y\"]]\n",
    "    * predict_df[config[\"column_names\"][\"y_hat\"]]\n",
    ").sum(axis=1).cumsum().plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3d38a3ec",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-06-07T15:16:51.217880Z",
     "start_time": "2022-06-07T15:16:51.176565Z"
    }
   },
   "source": [
    "# Get data for analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "443bc51a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:57.667810Z",
     "start_time": "2022-07-07T13:26:56.345584Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  removed rows with nans: 44027 / 887113 = 4.96%\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>volume</th>\n",
       "      <th>vwap</th>\n",
       "      <th>vwap.ret_0</th>\n",
       "      <th>vwap.ret_0.vol_adj</th>\n",
       "      <th>vwap.ret_0.vol_adj.c</th>\n",
       "      <th>vwap.ret_0.vol_adj.shift_-2</th>\n",
       "      <th>vwap.ret_0.vol_adj.shift_-2_hat</th>\n",
       "      <th>hit</th>\n",
       "      <th>trade_pnl</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>end_ts</th>\n",
       "      <th>asset_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">2018-01-04 12:00:00-05:00</th>\n",
       "      <th>binance::ETH_USDT</th>\n",
       "      <td>677.267720</td>\n",
       "      <td>993.835677</td>\n",
       "      <td>0.001891</td>\n",
       "      <td>1.954649</td>\n",
       "      <td>1.812616</td>\n",
       "      <td>-0.393211</td>\n",
       "      <td>0.075405</td>\n",
       "      <td>-1</td>\n",
       "      <td>-0.029650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>binance::BTC_USDT</th>\n",
       "      <td>117.433287</td>\n",
       "      <td>14469.739034</td>\n",
       "      <td>0.003955</td>\n",
       "      <td>1.437304</td>\n",
       "      <td>1.378481</td>\n",
       "      <td>1.259730</td>\n",
       "      <td>0.068264</td>\n",
       "      <td>1</td>\n",
       "      <td>0.085995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">2018-01-04 12:05:00-05:00</th>\n",
       "      <th>binance::ETH_USDT</th>\n",
       "      <td>936.989460</td>\n",
       "      <td>993.949461</td>\n",
       "      <td>0.000114</td>\n",
       "      <td>0.078415</td>\n",
       "      <td>0.078405</td>\n",
       "      <td>0.110683</td>\n",
       "      <td>0.087400</td>\n",
       "      <td>1</td>\n",
       "      <td>0.009674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>binance::BTC_USDT</th>\n",
       "      <td>77.354675</td>\n",
       "      <td>14432.219854</td>\n",
       "      <td>-0.002596</td>\n",
       "      <td>-1.253307</td>\n",
       "      <td>-1.213842</td>\n",
       "      <td>0.535206</td>\n",
       "      <td>-0.040640</td>\n",
       "      <td>-1</td>\n",
       "      <td>-0.021751</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-04 12:10:00-05:00</th>\n",
       "      <th>binance::ETH_USDT</th>\n",
       "      <td>655.104630</td>\n",
       "      <td>993.344329</td>\n",
       "      <td>-0.000609</td>\n",
       "      <td>-0.393211</td>\n",
       "      <td>-0.391949</td>\n",
       "      <td>0.642775</td>\n",
       "      <td>0.093588</td>\n",
       "      <td>1</td>\n",
       "      <td>0.060156</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 volume          vwap  vwap.ret_0  vwap.ret_0.vol_adj  vwap.ret_0.vol_adj.c  vwap.ret_0.vol_adj.shift_-2  vwap.ret_0.vol_adj.shift_-2_hat  hit  trade_pnl\n",
       "end_ts                    asset_id                                                                                                                                                         \n",
       "2018-01-04 12:00:00-05:00 binance::ETH_USDT  677.267720    993.835677    0.001891            1.954649              1.812616             -0.393211                  0.075405   -1  -0.029650\n",
       "                          binance::BTC_USDT  117.433287  14469.739034    0.003955            1.437304              1.378481              1.259730                  0.068264    1   0.085995\n",
       "2018-01-04 12:05:00-05:00 binance::ETH_USDT  936.989460    993.949461    0.000114            0.078415              0.078405              0.110683                  0.087400    1   0.009674\n",
       "                          binance::BTC_USDT   77.354675  14432.219854   -0.002596           -1.253307             -1.213842              0.535206                 -0.040640   -1  -0.021751\n",
       "2018-01-04 12:10:00-05:00 binance::ETH_USDT  655.104630    993.344329   -0.000609           -0.393211             -0.391949              0.642775                  0.093588    1   0.060156"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics_df = preprocess_predictions_df(config, predict_df)\n",
    "metrics_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "2016ceb0",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:57.912489Z",
     "start_time": "2022-07-07T13:26:57.670847Z"
    }
   },
   "outputs": [],
   "source": [
    "# Reset index to ease further preprocessing.\n",
    "# TODO(Dan): Move index resetting under plotting funtions.\n",
    "metrics_df_reset_index = metrics_df.reset_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "88f725d6",
   "metadata": {},
   "source": [
    "# Stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "f1876d33",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:57.979106Z",
     "start_time": "2022-07-07T13:26:57.916125Z"
    }
   },
   "outputs": [],
   "source": [
    "# Set oftenly used config parameters.\n",
    "asset_id = config[\"column_names\"][\"asset_id\"]\n",
    "timestamp = config[\"column_names\"][\"timestamp\"]\n",
    "volume = config[\"column_names\"][\"volume\"]\n",
    "y = config[\"column_names\"][\"y\"]\n",
    "y_hat = config[\"column_names\"][\"y_hat\"]\n",
    "hit = config[\"column_names\"][\"hit\"]\n",
    "trade_pnl = config[\"column_names\"][\"trade_pnl\"]\n",
    "#\n",
    "quantile_ranks = config[\"stats_kwargs\"][\"quantile_ranks\"]\n",
    "y_min_lim_hit_rate = config[\"plot_kwargs\"][\"y_min_lim_hit_rate\"]\n",
    "y_max_lim_hit_rate = config[\"plot_kwargs\"][\"y_max_lim_hit_rate\"]\n",
    "color = config[\"plot_kwargs\"][\"color\"]\n",
    "capsize = config[\"plot_kwargs\"][\"capsize\"]\n",
    "xticks_rotation = config[\"plot_kwargs\"][\"xticks_rotation\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9adc0d8",
   "metadata": {},
   "source": [
    "## By asset"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bef6de58",
   "metadata": {},
   "source": [
    "### Hit rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "394b5c73",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:58.784918Z",
     "start_time": "2022-07-07T13:26:57.982202Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2cb197786cba49838ab6fa2a545ea9ba",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hit_by_asset = calculate_hit_rate_with_CI(metrics_df_reset_index, asset_id, hit)\n",
    "plot_bars_with_widget(\n",
    "    hit_by_asset, \"hit_rate\", y_min_lim_hit_rate, y_max_lim_hit_rate\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "00479511",
   "metadata": {},
   "source": [
    "### PnL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "9437678a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:26:59.408486Z",
     "start_time": "2022-07-07T13:26:58.786563Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8ae69c4083b245f7b81b3d91b4d6603a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute PnL for each asset id.\n",
    "pnl_stats = (\n",
    "    metrics_df.groupby(asset_id)[trade_pnl].sum().sort_values(ascending=False)\n",
    ")\n",
    "pnl_stats = pnl_stats.rename(\"y\").to_frame()\n",
    "# Confidence Intervals are currently excluded.\n",
    "pnl_stats[\"errors\"] = 0\n",
    "# Plot PnL per asset id.\n",
    "# TODO(Max): infer y-limits automatically.\n",
    "plot_bars_with_widget(pnl_stats, \"avg_pnl_by_asset\", 0, 450)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "d93b8335",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:27:30.881450Z",
     "start_time": "2022-07-07T13:26:59.410958Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot cumulative PnL over time per asset id.\n",
    "_ = metrics_df[trade_pnl].dropna().unstack().cumsum().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "8eb14519",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:27:31.854526Z",
     "start_time": "2022-07-07T13:27:30.883945Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2a262cf6736144bfb1a943c46ee2eca4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot average trade PnL per asset id.\n",
    "avg_pnl_by_asset = calculate_CI_for_PnLs_or_SR(\n",
    "    metrics_df_reset_index, asset_id, trade_pnl\n",
    ")\n",
    "# TODO(Max): infer y-limits automatically.\n",
    "plot_bars_with_widget(avg_pnl_by_asset, \"avg_pnl_by_asset\", 0, 0.005)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e02e918e",
   "metadata": {},
   "source": [
    "### Sharpe Ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "347274ba",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:28:18.768594Z",
     "start_time": "2022-07-07T13:27:31.857911Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3e37fdde57614a39bc7effa4b7d162f0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute bootstrapped Sharpe Ratio.\n",
    "sr_by_asset = compute_sharpe_ratio(config, metrics_df_reset_index, asset_id)\n",
    "# Add CIs and errors.\n",
    "sr_ci_by_asset = calculate_CI_for_PnLs_or_SR(\n",
    "    sr_by_asset, asset_id, \"sharpe_ratio\"\n",
    ")\n",
    "# Visualize results.\n",
    "plot_bars_with_widget(sr_ci_by_asset, \"sharpe_ratio\", 0, 9)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ebb2ad71",
   "metadata": {},
   "source": [
    "## By time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "2406cf0e",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:28:19.685082Z",
     "start_time": "2022-07-07T13:28:18.773059Z"
    }
   },
   "outputs": [],
   "source": [
    "metrics_df_reset_index[\"hour\"] = metrics_df_reset_index[timestamp].dt.hour\n",
    "metrics_df_reset_index[\"weekday\"] = metrics_df_reset_index[\n",
    "    timestamp\n",
    "].dt.day_name()\n",
    "metrics_df_reset_index[\"month\"] = metrics_df_reset_index[\n",
    "    timestamp\n",
    "].dt.month_name()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4de2397a",
   "metadata": {},
   "source": [
    "### Hit Rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b458aed6",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:28:20.054787Z",
     "start_time": "2022-07-07T13:28:19.693005Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "547b605f3a34437daf8673d03ab09cfa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hits_by_time_hour = calculate_hit_rate_with_CI(\n",
    "    metrics_df_reset_index, \"hour\", hit\n",
    ")\n",
    "plot_bars_with_widget(\n",
    "    hits_by_time_hour, \"avg_hit_rate\", y_min_lim_hit_rate, y_max_lim_hit_rate\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "4b70798e",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:28:20.461346Z",
     "start_time": "2022-07-07T13:28:20.056610Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "73db6aeec817437c84ffc48adb49cc91",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hits_by_time_weekday = calculate_hit_rate_with_CI(\n",
    "    metrics_df_reset_index, \"weekday\", hit\n",
    ")\n",
    "plot_bars_with_widget(\n",
    "    hits_by_time_weekday, \"avg_hit_rate\", y_min_lim_hit_rate, y_max_lim_hit_rate\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "be5764e4",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:28:20.935158Z",
     "start_time": "2022-07-07T13:28:20.463207Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e7590de875f847e3939b4606bd872818",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hits_by_time_month = calculate_hit_rate_with_CI(\n",
    "    metrics_df_reset_index, \"month\", hit\n",
    ")\n",
    "plot_bars_with_widget(\n",
    "    hits_by_time_month, \"avg_hit_rate\", y_min_lim_hit_rate, y_max_lim_hit_rate\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1ba14ac",
   "metadata": {},
   "source": [
    "### PnL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "bf379b37",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:28:21.287641Z",
     "start_time": "2022-07-07T13:28:20.937706Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d1246eb6d1bf42dc80255a61a7e8d490",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pnl_by_time_hour = calculate_CI_for_PnLs_or_SR(\n",
    "    metrics_df_reset_index, \"hour\", trade_pnl\n",
    ")\n",
    "# TODO(Max): infer y-limits automatically.\n",
    "plot_bars_with_widget(pnl_by_time_hour, \"avg_pnl\", 0, 0.005)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "3b4e7353",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:28:21.704809Z",
     "start_time": "2022-07-07T13:28:21.289754Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "74ff584d8fa544f5ad40baaae47fdd84",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pnl_by_time_weekday = calculate_CI_for_PnLs_or_SR(\n",
    "    metrics_df_reset_index, \"weekday\", trade_pnl\n",
    ")\n",
    "# TODO(Max): infer y-limits automatically.\n",
    "plot_bars_with_widget(pnl_by_time_weekday, \"avg_pnl\", 0, 0.004)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "2976b08a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:28:22.156479Z",
     "start_time": "2022-07-07T13:28:21.707541Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "85fc5d16109d474bb354d48822d48a55",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pnl_by_time_month = calculate_CI_for_PnLs_or_SR(\n",
    "    metrics_df_reset_index, \"month\", trade_pnl\n",
    ")\n",
    "# TODO(Max): infer y-limits automatically.\n",
    "plot_bars_with_widget(pnl_by_time_month, \"avg_pnl\", 0, 0.0055)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f2199fad",
   "metadata": {},
   "source": [
    "### Sharpe Ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "5addec55",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:29:06.253658Z",
     "start_time": "2022-07-07T13:28:22.158476Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4d6a65f6535f41e98a840002c5592a08",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute bootstrapped Sharpe Ratio.\n",
    "sr_by_hour = compute_sharpe_ratio(config, metrics_df_reset_index, \"hour\")\n",
    "# Add CIs and errors.\n",
    "sr_ci_by_hour = calculate_CI_for_PnLs_or_SR(sr_by_hour, \"hour\", \"sharpe_ratio\")\n",
    "# Visualize results.\n",
    "plot_bars_with_widget(sr_ci_by_hour, \"sharpe_ratio\", 0, 9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "31bf4d54",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:29:36.158086Z",
     "start_time": "2022-07-07T13:29:06.256044Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "67e5bdd0a9764360a38938993fd4fdbe",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute bootstrapped Sharpe Ratio.\n",
    "sr_by_weekday = compute_sharpe_ratio(config, metrics_df_reset_index, \"weekday\")\n",
    "# Add CIs and errors.\n",
    "sr_ci_by_weekday = calculate_CI_for_PnLs_or_SR(\n",
    "    sr_by_weekday, \"weekday\", \"sharpe_ratio\"\n",
    ")\n",
    "# Visualize results.\n",
    "plot_bars_with_widget(sr_ci_by_weekday, \"sharpe_ratio\", 0, 7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "a09b5acd",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:30:13.690775Z",
     "start_time": "2022-07-07T13:29:36.161683Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cdee7bd8ec09465da5cd304fe8c8fa1d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute bootstrapped Sharpe Ratio.\n",
    "sr_by_month = compute_sharpe_ratio(config, metrics_df_reset_index, \"month\")\n",
    "# Add CIs and errors.\n",
    "sr_ci_by_month = calculate_CI_for_PnLs_or_SR(sr_by_month, \"month\", \"sharpe_ratio\")\n",
    "# Visualize results.\n",
    "plot_bars_with_widget(sr_ci_by_month, \"sharpe_ratio\", 0, 9)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c45a76a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-06-14T11:44:52.359488Z",
     "start_time": "2022-06-14T11:44:52.304630Z"
    }
   },
   "source": [
    "## By prediction magnitude"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "ca68ee01",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:30:13.856054Z",
     "start_time": "2022-07-07T13:30:13.708169Z"
    }
   },
   "outputs": [],
   "source": [
    "prediction_magnitude = \".\".join([y_hat, \"quantile_rank\"])\n",
    "metrics_df_reset_index[prediction_magnitude] = pd.qcut(\n",
    "    metrics_df_reset_index[y_hat], quantile_ranks, labels=False\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ef087c7",
   "metadata": {},
   "source": [
    "### Hit rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "c8b97960",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:30:14.411945Z",
     "start_time": "2022-07-07T13:30:13.858854Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "415e3b83c20e420d9bb23142baea8c42",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hits_by_prediction_magnitude = calculate_hit_rate_with_CI(\n",
    "    metrics_df_reset_index, prediction_magnitude, hit\n",
    ")\n",
    "plot_bars_with_widget(\n",
    "    hits_by_prediction_magnitude,\n",
    "    \"avg_hit_rate\",\n",
    "    y_min_lim_hit_rate,\n",
    "    y_max_lim_hit_rate,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fbe1e578",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-06-14T11:50:54.479960Z",
     "start_time": "2022-06-14T11:50:54.436323Z"
    }
   },
   "source": [
    "### PnL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "311cd7f9",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:30:14.792920Z",
     "start_time": "2022-07-07T13:30:14.413838Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4fc484dac88247c5a555b20d12f481be",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pnl_by_prediction_magnitude = calculate_CI_for_PnLs_or_SR(\n",
    "    metrics_df_reset_index, prediction_magnitude, trade_pnl\n",
    ")\n",
    "# TODO(Max): infer y-limits automatically.\n",
    "plot_bars_with_widget(pnl_by_prediction_magnitude, \"avg_pnl\", 0, 0.01)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "82a47a31",
   "metadata": {},
   "source": [
    "### Sharpe Ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "505a805f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:30:50.444990Z",
     "start_time": "2022-07-07T13:30:14.794601Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6608f6d17ce546839910303ddfe161c1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute bootstrapped Sharpe Ratio.\n",
    "sr_by_prediction_magnitude = compute_sharpe_ratio(\n",
    "    config, metrics_df_reset_index, prediction_magnitude\n",
    ")\n",
    "# Add CIs and errors.\n",
    "sr_ci_by_prediction_magnitude = calculate_CI_for_PnLs_or_SR(\n",
    "    sr_by_prediction_magnitude, prediction_magnitude, \"sharpe_ratio\"\n",
    ")\n",
    "# Visualize results.\n",
    "plot_bars_with_widget(sr_ci_by_prediction_magnitude, \"sharpe_ratio\", -1, 12)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b689693",
   "metadata": {},
   "source": [
    "## By volume"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "eecade72",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:30:50.792046Z",
     "start_time": "2022-07-07T13:30:50.446942Z"
    }
   },
   "outputs": [],
   "source": [
    "volume_quantile = \".\".join([volume, \"quantile_rank\"])\n",
    "metrics_df_reset_index[volume_quantile] = metrics_df_reset_index.groupby(\n",
    "    asset_id\n",
    ")[volume].transform(lambda x: pd.qcut(x, quantile_ranks, labels=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6c7b2ad",
   "metadata": {},
   "source": [
    "### Hit rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "3de4c8d0",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:30:51.222984Z",
     "start_time": "2022-07-07T13:30:50.794013Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c119f8fe1efc455cabd2042159a8e192",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hits_by_volume = calculate_hit_rate_with_CI(\n",
    "    metrics_df_reset_index, volume_quantile, hit\n",
    ")\n",
    "plot_bars_with_widget(\n",
    "    hits_by_volume, \"avg_hit_rate\", y_min_lim_hit_rate, y_max_lim_hit_rate\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "34dc2c05",
   "metadata": {},
   "source": [
    "### PnL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "dc998b0d",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:30:51.614993Z",
     "start_time": "2022-07-07T13:30:51.224845Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b8ec5101e894471a88150438920186c3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pnl_by_volume = calculate_CI_for_PnLs_or_SR(\n",
    "    metrics_df_reset_index, volume_quantile, trade_pnl\n",
    ")\n",
    "# TODO(Max): infer y-limits automatically.\n",
    "plot_bars_with_widget(pnl_by_volume, \"avg_pnl\", 0.0005, 0.0045)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "763d33da",
   "metadata": {},
   "source": [
    "### Sharpe Ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "100442f0",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-07-07T13:31:28.360569Z",
     "start_time": "2022-07-07T13:30:51.617441Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0354da4e314746d7bfa62838e9fac9c2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(ToggleButtons(description='Sort by:', options=('x', 'y', 'ci_low', 'ci_high', False), vaÃÂÃÂ¢ÃÂÃÂÃÂÃÂ¦"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute bootstrapped Sharpe Ratio.\n",
    "sr_by_volume_quantile = compute_sharpe_ratio(\n",
    "    config, metrics_df_reset_index, volume_quantile\n",
    ")\n",
    "# Add CIs and errors.\n",
    "sr_ci_by_volume_quantile = calculate_CI_for_PnLs_or_SR(\n",
    "    sr_by_volume_quantile, volume_quantile, \"sharpe_ratio\"\n",
    ")\n",
    "# Visualize results.\n",
    "plot_bars_with_widget(sr_ci_by_volume_quantile, \"sharpe_ratio\", 0, 9)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.10"
  },
  "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": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "165px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}