dataflow/model/notebooks/Master_model_analyzer.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "a0cfeac2",
   "metadata": {},
   "source": [
    "# Description\n",
    "\n",
    "- Initialize with returns, predictions, target volatility, and oos start date\n",
    "- Evaluate portfolios generated from the predictions\n",
    "\n",
    "- TODO(gp): This should be called `Master_model_evaluator` like the class"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "126c8435",
   "metadata": {},
   "source": [
    "# Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "abf41299",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:39.095791Z",
     "start_time": "2022-02-23T13:02:33.959156Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/app/amp/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"
     ]
    }
   ],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "import logging\n",
    "\n",
    "import core.config as cconfig\n",
    "import dataflow.model.model_evaluator as dtfmomoeva\n",
    "import dataflow.model.model_plotter as dtfmomoplo\n",
    "import helpers.hdbg as hdbg\n",
    "import helpers.hprint as hprint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "eee79ad8",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:39.173834Z",
     "start_time": "2022-02-23T13:02:39.099168Z"
    }
   },
   "outputs": [
    {
     "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-d4f1f855-016a-4054-be38-710287312fd3.json'\n",
      "WARNING: Running in Jupyter\n",
      "\u001b[33mWARNING\u001b[0m: Disabling annoying warnings\n"
     ]
    }
   ],
   "source": [
    "hdbg.init_logger(verbosity=logging.INFO)\n",
    "# hdbg.init_logger(verbosity=logging.DEBUG)\n",
    "\n",
    "_LOG = logging.getLogger(__name__)\n",
    "\n",
    "# _LOG.info(\"%s\", env.get_system_signature()[0])\n",
    "\n",
    "hprint.config_notebook()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c0c17e86",
   "metadata": {},
   "source": [
    "# Notebook config"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a9c3584e",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:39.261499Z",
     "start_time": "2022-02-23T13:02:39.177781Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING Environment variable 'AM_CONFIG_CODE' not defined: no config retrieved\n",
      "load_experiment_kwargs:\n",
      "  src_dir: /app/experiment.RH1E.ccxt_v1-top2.5T.2018_2022\n",
      "  file_name: result_bundle.v2_0.pkl\n",
      "  experiment_type: ins_oos\n",
      "  selected_idxs: None\n",
      "  aws_profile: None\n",
      "model_evaluator_kwargs:\n",
      "  predictions_col: ret_0_vol_adj_2_hat\n",
      "  target_col: ret_0_vol_adj_2\n",
      "  oos_start: None\n",
      "  abort_on_error: True\n",
      "bh_adj_threshold: 0.1\n",
      "resample_rule: W\n",
      "mode: ins\n",
      "target_volatility: 0.1\n"
     ]
    }
   ],
   "source": [
    "# Read from env var.\n",
    "eval_config = cconfig.Config.from_env_var(\"AM_CONFIG_CODE\")\n",
    "\n",
    "# Override config.\n",
    "if eval_config is None:\n",
    "    # experiment_dir = \"/cache/experiments/oos_experiment.RH1E.v2_0-top100.5T\"\n",
    "    # experiment_dir = \"/app/rc_experiment.RH8Ec.v2_0-top2.5T.2009.run1\"\n",
    "    # experiment_dir = \"/app/experiment.RH1E.kibot_v2-top2.5T\"\n",
    "    experiment_dir = \"/app/experiment.RH1E.ccxt_v1-top2.5T.2018_2022\"\n",
    "    # experiment_dir = \"/app/experiment.RH1E.kibot_v2-top20.5T\"\n",
    "    aws_profile = None\n",
    "    selected_idxs = None\n",
    "\n",
    "    eval_config = cconfig.Config.from_dict(\n",
    "        {\n",
    "            \"load_experiment_kwargs\": {\n",
    "                \"src_dir\": experiment_dir,\n",
    "                \"file_name\": \"result_bundle.v2_0.pkl\",\n",
    "                \"experiment_type\": \"ins_oos\",\n",
    "                \"selected_idxs\": selected_idxs,\n",
    "                \"aws_profile\": aws_profile,\n",
    "            },\n",
    "            \"model_evaluator_kwargs\": {\n",
    "                # \"predictions_col\": \"mid_ret_0_vol_adj_clipped_2_hat\",\n",
    "                # \"target_col\": \"mid_ret_0_vol_adj_clipped_2\",\n",
    "                \"predictions_col\": \"ret_0_vol_adj_2_hat\",\n",
    "                \"target_col\": \"ret_0_vol_adj_2\",\n",
    "                # \"oos_start\": \"2017-01-01\",\n",
    "                \"oos_start\": None,\n",
    "                \"abort_on_error\": True,\n",
    "            },\n",
    "            \"bh_adj_threshold\": 0.1,\n",
    "            \"resample_rule\": \"W\",\n",
    "            \"mode\": \"ins\",\n",
    "            \"target_volatility\": 0.1,\n",
    "        }\n",
    "    )\n",
    "\n",
    "print(str(eval_config))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8142e65d",
   "metadata": {},
   "source": [
    "# Initialize ModelEvaluator and ModelPlotter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "371704c7",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:40.183469Z",
     "start_time": "2022-02-23T13:02:39.265626Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  Before load_experiment_artifacts: memory_usage=rss=0.301GB vms=2.870GB mem_pct=3%\n",
      "INFO  # Load artifacts 'result_bundle.v2_0.pkl' from '/app/experiment.RH1E.ccxt_v1-top2.5T.2018_2022'\n",
      "INFO  Found 2 experiment subdirs in '/app/experiment.RH1E.ccxt_v1-top2.5T.2018_2022'\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bb90e3441a97451b8a6388a549106593",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Loading artifacts:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  load_experiment_artifacts: memory_usage=rss=0.343GB vms=3.496GB mem_pct=3%\n",
      "INFO  load_experiment_artifacts: memory_usage=rss=0.352GB vms=3.504GB mem_pct=3%\n",
      "INFO  After load_experiment_artifacts: memory_usage=rss=0.352GB vms=3.504GB mem_pct=3%\n",
      "INFO  Before building ModelEvaluator: memory_usage=rss=0.352GB vms=3.504GB mem_pct=3%\n",
      "INFO  After building ModelEvaluator: memory_usage=rss=0.358GB vms=3.510GB mem_pct=3%\n"
     ]
    }
   ],
   "source": [
    "# Build the ModelEvaluator from the eval config.\n",
    "evaluator = dtfmomoeva.ModelEvaluator.from_eval_config(eval_config)\n",
    "\n",
    "# Build the ModelPlotter.\n",
    "plotter = dtfmomoplo.ModelPlotter(evaluator)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2ab68b06",
   "metadata": {},
   "source": [
    "# Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "f353ea8a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:42.691500Z",
     "start_time": "2022-02-23T13:02:40.185350Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1866d1beef2c4f6b9b84175c558f49d4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating positions:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0a42d870447f4e7ba7e7639c4c79b961",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating PnL:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  memory_usage=rss=0.398GB vms=3.549GB mem_pct=3%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e85528ae55184ac88c02ed0cb8660930",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating stats:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  stats=\n",
      "avg_turnover_(%)           143.533932\n",
      "turnover_frequency      <5 * Minutes>\n",
      "avg_holding_period           0.696699\n",
      "holding_period_units    <5 * Minutes>\n",
      "Name: positions, dtype: object\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/venv/lib/python3.8/site-packages/statsmodels/tsa/stattools.py:2015: InterpolationWarning: The test statistic is outside of the range of p-values available in the\n",
      "look-up table. The actual p-value is greater than the p-value returned.\n",
      "\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  stats=\n",
      "avg_turnover_(%)           151.108831\n",
      "turnover_frequency      <5 * Minutes>\n",
      "avg_holding_period           0.661775\n",
      "holding_period_units    <5 * Minutes>\n",
      "Name: positions, dtype: object\n",
      "INFO  memory_usage=rss=0.380GB vms=3.628GB mem_pct=3%\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/venv/lib/python3.8/site-packages/statsmodels/tsa/stattools.py:2011: InterpolationWarning: The test statistic is outside of the range of p-values available in the\n",
      "look-up table. The actual p-value is smaller than the p-value returned.\n",
      "\n",
      "  warnings.warn(\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>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">portfolio</th>\n",
       "      <th>avg_turnover_(%)</th>\n",
       "      <td>143.533932</td>\n",
       "      <td>151.108831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>turnover_frequency</th>\n",
       "      <td>&lt;5 * Minutes&gt;</td>\n",
       "      <td>&lt;5 * Minutes&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>avg_holding_period</th>\n",
       "      <td>0.696699</td>\n",
       "      <td>0.661775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>holding_period_units</th>\n",
       "      <td>&lt;5 * Minutes&gt;</td>\n",
       "      <td>&lt;5 * Minutes&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">ratios</th>\n",
       "      <th>sharpe_ratio</th>\n",
       "      <td>1.781133</td>\n",
       "      <td>1.441202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sharpe_ratio_standard_error</th>\n",
       "      <td>0.535551</td>\n",
       "      <td>0.849651</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sr.tval</th>\n",
       "      <td>3.326076</td>\n",
       "      <td>1.696323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sr.pval</th>\n",
       "      <td>0.000881</td>\n",
       "      <td>0.089842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>kratio</th>\n",
       "      <td>2.610733</td>\n",
       "      <td>1.268196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"18\" valign=\"top\">sampling</th>\n",
       "      <th>start_time</th>\n",
       "      <td>2019-01-15 10:20:00-05:00</td>\n",
       "      <td>2021-03-02 15:20:00-05:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>end_time</th>\n",
       "      <td>2022-02-09 15:50:00-05:00</td>\n",
       "      <td>2022-02-09 15:50:00-05:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_sampling_points</th>\n",
       "      <td>56650</td>\n",
       "      <td>17719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>frequency</th>\n",
       "      <td>&lt;5 * Minutes&gt;</td>\n",
       "      <td>&lt;5 * Minutes&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sampling_points_per_year</th>\n",
       "      <td>105156.125</td>\n",
       "      <td>105156.125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time_span_in_years</th>\n",
       "      <td>3.070815</td>\n",
       "      <td>0.942209</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_rows</th>\n",
       "      <td>366641</td>\n",
       "      <td>145667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>frac_zero</th>\n",
       "      <td>0.006156</td>\n",
       "      <td>0.001181</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>frac_nan</th>\n",
       "      <td>0.845489</td>\n",
       "      <td>0.87836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>frac_inf</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>frac_constant</th>\n",
       "      <td>0.000412</td>\n",
       "      <td>0.000034</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_finite_samples</th>\n",
       "      <td>56650</td>\n",
       "      <td>17719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_finite_samples_inv</th>\n",
       "      <td>0.000018</td>\n",
       "      <td>0.000056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_finite_samples_inv_dyadic_scale</th>\n",
       "      <td>-16</td>\n",
       "      <td>-15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_finite_samples_sqrt</th>\n",
       "      <td>238.012605</td>\n",
       "      <td>133.112734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_finite_samples_sqrt_inv</th>\n",
       "      <td>0.004201</td>\n",
       "      <td>0.007512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_finite_samples_sqrt_inv_dyadic_scale</th>\n",
       "      <td>-8</td>\n",
       "      <td>-8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_unique_values</th>\n",
       "      <td>54394</td>\n",
       "      <td>17548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"15\" valign=\"top\">summary</th>\n",
       "      <th>scipy.mean</th>\n",
       "      <td>0.000011</td>\n",
       "      <td>0.000011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>scipy.std</th>\n",
       "      <td>0.000784</td>\n",
       "      <td>0.000884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>scipy.skew</th>\n",
       "      <td>0.247449</td>\n",
       "      <td>0.115654</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>scipy.kurtosis</th>\n",
       "      <td>103.762631</td>\n",
       "      <td>158.445509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>null_mean_zero.tval</th>\n",
       "      <td>3.326076</td>\n",
       "      <td>1.696323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>null_mean_zero.pval</th>\n",
       "      <td>0.000881</td>\n",
       "      <td>0.089842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>jensen_ratio</th>\n",
       "      <td>0.34411</td>\n",
       "      <td>0.364232</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>56650.0</td>\n",
       "      <td>17719.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.000011</td>\n",
       "      <td>0.000011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.000784</td>\n",
       "      <td>0.000884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-0.017151</td>\n",
       "      <td>-0.023424</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.000067</td>\n",
       "      <td>-0.000095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.000077</td>\n",
       "      <td>0.000106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>0.017309</td>\n",
       "      <td>0.023424</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"14\" valign=\"top\">stationarity</th>\n",
       "      <th>adf.stat</th>\n",
       "      <td>-66.767514</td>\n",
       "      <td>-31.527366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adf.pval</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adf.used_lag</th>\n",
       "      <td>14.0</td>\n",
       "      <td>16.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adf.nobs</th>\n",
       "      <td>56635.0</td>\n",
       "      <td>17702.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adf.critical_values_1%</th>\n",
       "      <td>-3.430465</td>\n",
       "      <td>-3.430719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adf.critical_values_5%</th>\n",
       "      <td>-2.861591</td>\n",
       "      <td>-2.861703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adf.critical_values_10%</th>\n",
       "      <td>-2.566797</td>\n",
       "      <td>-2.566857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>adf.ic_best</th>\n",
       "      <td>-649596.640168</td>\n",
       "      <td>-198780.438983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>kpss.stat</th>\n",
       "      <td>0.079126</td>\n",
       "      <td>0.834879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>kpss.pval</th>\n",
       "      <td>0.1</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>kpss.lags</th>\n",
       "      <td>16.0</td>\n",
       "      <td>16.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>kpss.critical_values_1%</th>\n",
       "      <td>0.739</td>\n",
       "      <td>0.739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>kpss.critical_values_5%</th>\n",
       "      <td>0.463</td>\n",
       "      <td>0.463</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>kpss.critical_values_10%</th>\n",
       "      <td>0.347</td>\n",
       "      <td>0.347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">normality</th>\n",
       "      <th>omnibus_null_normal.stat</th>\n",
       "      <td>22796.259802</td>\n",
       "      <td>7578.999834</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>omnibus_null_normal.pval</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>centered_gaussian.log_likelihood</th>\n",
       "      <td>324688.79316</td>\n",
       "      <td>99437.220462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>centered_gaussian.centered_var</th>\n",
       "      <td>0.000001</td>\n",
       "      <td>0.000001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>spectral</th>\n",
       "      <th>forecastability</th>\n",
       "      <td>0.003691</td>\n",
       "      <td>0.003338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">portfolio</th>\n",
       "      <th>annualized_mean_return</th>\n",
       "      <td>0.178113</td>\n",
       "      <td>0.14412</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>annualized_volatility</th>\n",
       "      <td>0.1</td>\n",
       "      <td>0.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max_drawdown</th>\n",
       "      <td>0.061997</td>\n",
       "      <td>0.144937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hit_rate_point_est_(%)</th>\n",
       "      <td>50.824555</td>\n",
       "      <td>49.940161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hit_rate_97.50%CI_lower_bound_(%)</th>\n",
       "      <td>50.404386</td>\n",
       "      <td>49.200418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hit_rate_97.50%CI_upper_bound_(%)</th>\n",
       "      <td>51.244635</td>\n",
       "      <td>50.679923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>correlation</th>\n",
       "      <th>prediction_corr_implied_by_pnl</th>\n",
       "      <td>0.013975</td>\n",
       "      <td>0.012744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"13\" valign=\"top\">bets</th>\n",
       "      <th>num_positions</th>\n",
       "      <td>56638</td>\n",
       "      <td>17719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_bets</th>\n",
       "      <td>28884</td>\n",
       "      <td>9674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>long_bets_frac</th>\n",
       "      <td>0.499965</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_num_bets_per_year</th>\n",
       "      <td>8284.205843</td>\n",
       "      <td>6983.602005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_bet_length</th>\n",
       "      <td>1.960878</td>\n",
       "      <td>1.831611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bet_length_units</th>\n",
       "      <td>5 * Minutes</td>\n",
       "      <td>5 * Minutes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bet_hit_rate_point_est_(%)</th>\n",
       "      <td>51.420465</td>\n",
       "      <td>52.14473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bet_hit_rate_97.50%CI_lower_bound_(%)</th>\n",
       "      <td>50.83328</td>\n",
       "      <td>51.141537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bet_hit_rate_97.50%CI_upper_bound_(%)</th>\n",
       "      <td>52.007355</td>\n",
       "      <td>53.146621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_return_winning_bets</th>\n",
       "      <td>0.000346</td>\n",
       "      <td>0.000411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_return_losing_bets</th>\n",
       "      <td>-0.000355</td>\n",
       "      <td>-0.000434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_return_long_bet</th>\n",
       "      <td>0.000022</td>\n",
       "      <td>0.000023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_return_short_bet</th>\n",
       "      <td>-0.000012</td>\n",
       "      <td>-0.00001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>correlation</th>\n",
       "      <th>pnl_corr_to_underlying</th>\n",
       "      <td>0.033447</td>\n",
       "      <td>-0.00833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ratios</th>\n",
       "      <th>sr.adj_pval</th>\n",
       "      <td>0.001763</td>\n",
       "      <td>0.089842</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                               0                          1\n",
       "portfolio    avg_turnover_(%)                                         143.533932                 151.108831\n",
       "             turnover_frequency                                    <5 * Minutes>              <5 * Minutes>\n",
       "             avg_holding_period                                         0.696699                   0.661775\n",
       "             holding_period_units                                  <5 * Minutes>              <5 * Minutes>\n",
       "ratios       sharpe_ratio                                               1.781133                   1.441202\n",
       "             sharpe_ratio_standard_error                                0.535551                   0.849651\n",
       "             sr.tval                                                    3.326076                   1.696323\n",
       "             sr.pval                                                    0.000881                   0.089842\n",
       "             kratio                                                     2.610733                   1.268196\n",
       "sampling     start_time                                2019-01-15 10:20:00-05:00  2021-03-02 15:20:00-05:00\n",
       "             end_time                                  2022-02-09 15:50:00-05:00  2022-02-09 15:50:00-05:00\n",
       "             n_sampling_points                                             56650                      17719\n",
       "             frequency                                             <5 * Minutes>              <5 * Minutes>\n",
       "             sampling_points_per_year                                 105156.125                 105156.125\n",
       "             time_span_in_years                                         3.070815                   0.942209\n",
       "             n_rows                                                       366641                     145667\n",
       "             frac_zero                                                  0.006156                   0.001181\n",
       "             frac_nan                                                   0.845489                    0.87836\n",
       "             frac_inf                                                        0.0                        0.0\n",
       "             frac_constant                                              0.000412                   0.000034\n",
       "             num_finite_samples                                            56650                      17719\n",
       "             num_finite_samples_inv                                     0.000018                   0.000056\n",
       "             num_finite_samples_inv_dyadic_scale                             -16                        -15\n",
       "             num_finite_samples_sqrt                                  238.012605                 133.112734\n",
       "             num_finite_samples_sqrt_inv                                0.004201                   0.007512\n",
       "             num_finite_samples_sqrt_inv_dyadic_scale                         -8                         -8\n",
       "             num_unique_values                                             54394                      17548\n",
       "summary      scipy.mean                                                 0.000011                   0.000011\n",
       "             scipy.std                                                  0.000784                   0.000884\n",
       "             scipy.skew                                                 0.247449                   0.115654\n",
       "             scipy.kurtosis                                           103.762631                 158.445509\n",
       "             null_mean_zero.tval                                        3.326076                   1.696323\n",
       "             null_mean_zero.pval                                        0.000881                   0.089842\n",
       "             jensen_ratio                                                0.34411                   0.364232\n",
       "             count                                                       56650.0                    17719.0\n",
       "             mean                                                       0.000011                   0.000011\n",
       "             std                                                        0.000784                   0.000884\n",
       "             min                                                       -0.017151                  -0.023424\n",
       "             25%                                                       -0.000067                  -0.000095\n",
       "             50%                                                             0.0                        0.0\n",
       "             75%                                                        0.000077                   0.000106\n",
       "             max                                                        0.017309                   0.023424\n",
       "stationarity adf.stat                                                 -66.767514                 -31.527366\n",
       "             adf.pval                                                        0.0                        0.0\n",
       "             adf.used_lag                                                   14.0                       16.0\n",
       "             adf.nobs                                                    56635.0                    17702.0\n",
       "             adf.critical_values_1%                                    -3.430465                  -3.430719\n",
       "             adf.critical_values_5%                                    -2.861591                  -2.861703\n",
       "             adf.critical_values_10%                                   -2.566797                  -2.566857\n",
       "             adf.ic_best                                          -649596.640168             -198780.438983\n",
       "             kpss.stat                                                  0.079126                   0.834879\n",
       "             kpss.pval                                                       0.1                       0.01\n",
       "             kpss.lags                                                      16.0                       16.0\n",
       "             kpss.critical_values_1%                                       0.739                      0.739\n",
       "             kpss.critical_values_5%                                       0.463                      0.463\n",
       "             kpss.critical_values_10%                                      0.347                      0.347\n",
       "normality    omnibus_null_normal.stat                               22796.259802                7578.999834\n",
       "             omnibus_null_normal.pval                                        0.0                        0.0\n",
       "             centered_gaussian.log_likelihood                       324688.79316               99437.220462\n",
       "             centered_gaussian.centered_var                             0.000001                   0.000001\n",
       "spectral     forecastability                                            0.003691                   0.003338\n",
       "portfolio    annualized_mean_return                                     0.178113                    0.14412\n",
       "             annualized_volatility                                           0.1                        0.1\n",
       "             max_drawdown                                               0.061997                   0.144937\n",
       "             hit_rate_point_est_(%)                                    50.824555                  49.940161\n",
       "             hit_rate_97.50%CI_lower_bound_(%)                         50.404386                  49.200418\n",
       "             hit_rate_97.50%CI_upper_bound_(%)                         51.244635                  50.679923\n",
       "correlation  prediction_corr_implied_by_pnl                             0.013975                   0.012744\n",
       "bets         num_positions                                                 56638                      17719\n",
       "             num_bets                                                      28884                       9674\n",
       "             long_bets_frac                                             0.499965                        0.5\n",
       "             mean_num_bets_per_year                                  8284.205843                6983.602005\n",
       "             mean_bet_length                                            1.960878                   1.831611\n",
       "             bet_length_units                                        5 * Minutes                5 * Minutes\n",
       "             bet_hit_rate_point_est_(%)                                51.420465                   52.14473\n",
       "             bet_hit_rate_97.50%CI_lower_bound_(%)                      50.83328                  51.141537\n",
       "             bet_hit_rate_97.50%CI_upper_bound_(%)                     52.007355                  53.146621\n",
       "             mean_return_winning_bets                                   0.000346                   0.000411\n",
       "             mean_return_losing_bets                                   -0.000355                  -0.000434\n",
       "             mean_return_long_bet                                       0.000022                   0.000023\n",
       "             mean_return_short_bet                                     -0.000012                   -0.00001\n",
       "correlation  pnl_corr_to_underlying                                     0.033447                   -0.00833\n",
       "ratios       sr.adj_pval                                                0.001763                   0.089842"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pnl_stats = evaluator.calculate_stats(\n",
    "    mode=eval_config[\"mode\"], target_volatility=eval_config[\"target_volatility\"]\n",
    ")\n",
    "display(pnl_stats)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c36bf064",
   "metadata": {},
   "source": [
    "## Model selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b6fecfb5",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:43.348387Z",
     "start_time": "2022-02-23T13:02:42.698721Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cb9c1c0498874da08d9c1255fea1af75",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating positions:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a13ee080d2344bdcbce8f24a5dcc2e19",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating PnL:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  memory_usage=rss=0.396GB vms=3.644GB mem_pct=3%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotter.plot_multiple_tests_adjustment(\n",
    "    threshold=eval_config[\"bh_adj_threshold\"], mode=eval_config[\"mode\"]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "b5b44ae9",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:43.544688Z",
     "start_time": "2022-02-23T13:02:43.481666Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num model selected=2 / 2 = 100.00%\n",
      "model selected=[0, 1]\n",
      "model not selected=[]\n"
     ]
    }
   ],
   "source": [
    "# TODO(gp): Move this chunk of code in a function.\n",
    "col_mask = (\n",
    "    pnl_stats.loc[\"ratios\"].loc[\"sr.adj_pval\"] < eval_config[\"bh_adj_threshold\"]\n",
    ")\n",
    "selected = pnl_stats.loc[:, col_mask].columns.to_list()\n",
    "not_selected = pnl_stats.loc[:, ~col_mask].columns.to_list()\n",
    "\n",
    "print(\"num model selected=%s\" % hprint.perc(len(selected), pnl_stats.shape[1]))\n",
    "print(\"model selected=%s\" % selected)\n",
    "print(\"model not selected=%s\" % not_selected)\n",
    "\n",
    "# Use `selected = None` to show all the models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "1736ff88",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:45.923492Z",
     "start_time": "2022-02-23T13:02:43.548132Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ead347a21aa94084960c3f06b5b19912",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating positions:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c5fc783431514cf8bfc00310d0ba2bd8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating PnL:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  memory_usage=rss=0.387GB vms=3.635GB mem_pct=3%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5c900a98444b43cea85a130020f3f5c0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating positions:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "43b3b6df3bc14bfc9694e22189fd6ef8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating PnL:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  memory_usage=rss=0.402GB vms=3.650GB mem_pct=3%\n",
      "INFO  stats=\n",
      "avg_turnover_(%)            144.81124\n",
      "turnover_frequency      <5 * Minutes>\n",
      "avg_holding_period           0.690554\n",
      "holding_period_units    <5 * Minutes>\n",
      "Name: positions, dtype: object\n",
      "INFO  memory_usage=rss=0.422GB vms=3.669GB mem_pct=4%\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/venv/lib/python3.8/site-packages/statsmodels/tsa/stattools.py:2015: InterpolationWarning: The test statistic is outside of the range of p-values available in the\n",
      "look-up table. The actual p-value is greater than the p-value returned.\n",
      "\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# selected = None\n",
    "plotter.plot_multiple_pnls(\n",
    "    keys=selected,\n",
    "    resample_rule=eval_config[\"resample_rule\"],\n",
    "    mode=eval_config[\"mode\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9821d152",
   "metadata": {},
   "source": [
    "## Return correlation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "1d0698e1",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.332637Z",
     "start_time": "2022-02-23T13:02:45.925239Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "43618c28ffb4480e9a7d33ce02456590",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating positions:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "44422e98265e401bbc91bd90cb53c06a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Calculating PnL:   0%|          | 0/2 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO  memory_usage=rss=0.383GB vms=3.632GB mem_pct=3%\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAExCAYAAACd/QIBAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAQrUlEQVR4nO3dX2iUV4PH8V/m0UKpHXSCCU/+gG1p7fSP0CIt9E+gZuIoTqvDommnF0GKpWgtuxdb04smRr1oXGhxDb3aUlJiJQ2rax1tkqawLSp0eSksYgrablKNmSQ0gzs33uzM7MW+r4tMRuPJJDPH8/3AgHly5nnO1TfHM8/MVOXz+bwAAPe1QLknAABYfMQeABxA7AHAAcQeABxA7AHAAcQeABywrNwTuN9ceSVa7imgQj0YbS73FFDBGj7+cFHPz8oeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxQsth3d3drw4YNWrt2rS5fvjznmGw2q66uLkUiEbW0tGhgYKBUlwcAp4yNjam1tVXRaFStra0aHx+/4/iSxb65uVnHjh1TfX190TGnT5/W1atXNTw8rP7+fh09elQTExOlmgIAOKOzs1OJREJDQ0NKJBLq6Oi44/iSxX79+vXyff+OY86ePavt27crEAgoFAopEolocHCwYFwmk9HExETBI5PJlGq6AGCt2dlZjY6OKhaLSZJisZhGR0eVTqeLPmfZUk1OklKplOrq6m797Pu+pqamCsb19vaqp6en4Hhy9RpNFzn34+eGSjVNAKhoqVRKtbW18jxPkuR5nmpqapRKpRQKheZ8zpLGfr7a2toUj8cLjt98a1cZZgMA9lvS2Pu+r8nJSa1bt05S4Ur/b4LBoILBYMHxKx43DwGA7/uanp5WNpuV53nKZrOamZm541b6ktZz06ZNGhgYUC6XUzqd1sjIiKLR6LyfX+V5RR8A4Irq6mqFw2Elk0lJUjKZVDgcLrqFI5Uw9ocOHVJTU5Ompqa0c+dObdmyRZK0a9cuXbx4UZK0detWNTQ0aOPGjdqxY4f27NmjxsbG+V/E84o/AMAh+/fvV19fn6LRqPr6+tTV1XXH8VX5fD6/RHNbsN83/V3R3z02+K9LOJPirrwy//+pwC0PRpvLPQVUsIaPP1zU81fkC7RFsYIHACPEHgAcYFXsq7gbBwCMWBV7VvYAYMau2FdVlXsGAGAlq2LP/fQAYMaq2LONAwBmrIo9K3sAMGNV7MXdOABgxLLYs7IHABNWxb4qQOwBwIRVsWcbBwDMWBZ7VvYAYMKq2FcFeFMVAJiwKvZizx4AjNgVe7ZxAMCIVbHnUy8BwIxVsWdlDwBmrIo9H5cAAGasij332QOAGctiz8oeAEzYFXu+vAQAjFgVe/bsAcCMVbFnGwcAzFgVe+6zBwAzJYv92NiY2tvbdePGDa1cuVLd3d1as2bNbWOOHj2qr7/+WjU1NZKk559/Xp2dnfO/CB+XAABGShb7zs5OJRIJbd26VadOnVJHR4e++uqrgnHbtm3Tvn37zC7CNg4AGClJ7GdnZzU6Oqovv/xSkhSLxXTw4EGl02mFQqF7Pl8mk1Emkyk47rGNAwBGShL7VCql2tpaeX9deXuep5qaGqVSqYLYnzlzRufOndPq1au1d+9ePffccwXn6+3tVU9PT8Hxf/8Hw/8RAIDjlvQF2jfffFPvvfeeli9frvPnz2v37t06e/asVq1addu4trY2xePxgudXfTeyVFMFgPtKSWLv+76mp6eVzWbleZ6y2axmZmbk+/5t41avXn3r3y+//LJ839eVK1f0wgsv3DYuGAwqGAwWXGeqim0cADBRknpWV1crHA4rmUxKkpLJpMLhcMEWzvT09K1///rrr7p+/boeeeSR+V/ICxR/AACKKtk2zv79+9Xe3q7PP/9cwWBQ3d3dkqRdu3bpgw8+0LPPPqtPP/1Uly5dUiAQ0PLly3X48OHbVvt3wztoAcBMVT6fz5d7EvM1feybor+rfXvHEs6kuCuvRMs9BVSoB6PN5Z4CKljDxx8u6vmtegetAmzXAIAJq2LPNg4AmLEq9ryDFgDMWBV7PggNAMxYFXtxnz0AGLEr9qzsAcCIZbFnzx4ATFgVe+7GAQAzVsWebRwAMGNV7FnZA4AZq2LPnj0AmLEs9mzjAIAJq2JfxReOA4ARq2KvQFW5ZwAAVrIr9uzZA4ARq2LP3TgAYMaq2PN59gBgxqrYs7IHADNWxZ49ewAwY1ns2cYBABNWxZ5tHAAwY1Xs2cYBADN2xb6KN1UBgAmrYs82DgCYsSr2vEALAGZKVs+xsTG1trYqGo2qtbVV4+PjBWOy2ay6uroUiUTU0tKigYGBe7uI5xV/AACKKlnsOzs7lUgkNDQ0pEQioY6OjoIxp0+f1tWrVzU8PKz+/n4dPXpUExMT875GlecVfQAAiitJ7GdnZzU6OqpYLCZJisViGh0dVTqdvm3c2bNntX37dgUCAYVCIUUiEQ0ODhacL5PJaGJiouChQKD4AwBQVEn27FOplGpra+X9dYXteZ5qamqUSqUUCoVuG1dXV3frZ9/3NTU1VXC+3t5e9fT0FBz/j7/8pRTTXVQPRpvLPQVUqJtDP5R7CqhkH3+4qKevyBdo29raFI/HC47/Ty5fhtkAgP1KEnvf9zU9Pa1sNivP85TNZjUzMyPf9wvGTU5Oat26dZIKV/p/EwwGFQwGC47/eeO/SzFdAHBOSTa7q6urFQ6HlUwmJUnJZFLhcPi2LRxJ2rRpkwYGBpTL5ZROpzUyMqJoNDrv62Rz+aIPAEBxJXtlc//+/err61M0GlVfX5+6urokSbt27dLFixclSVu3blVDQ4M2btyoHTt2aM+ePWpsbJz3NXK5XNEHAKC4qnw+b82yeGJmtujvGmqql3AmxU0cPFzuKaBC8QIt7uTxc0OLev6KfIG2mGyeFTwAmLAr9uzNA4ARy2LPyh4ATFgV+xwrewAwYlXsWdkDgBmrYs+6HgDMWBV7VvYAYMay2LO2BwATlsWelT0AmLAs9qzsAcCEVbHP8Q5aADBiVexZ2QOAGctiz8oeAExYFXuLPqATACqKVbFnGwcAzNgVe16gBQAjVsWeD0IDADNWxZ4XaAHAjGWxZ2UPACYsiz0rewAwYVfsufUSAIxYFfscK3sAMGJX7FnYA4ARu2LPyh4AjFgVe+7GAQAzC479zZs39dFHH+nSpUvyPE/79u3Ta6+9VjDu559/1rvvvqs1a9ZIkh544AENDAzc07V4By0AmFlw7L/44gutWLFC33//vcbHx/X2229reHhYDz30UMHYxx57TCdOnDC+Fit7ADCz4Nh/9913+uSTTyRJa9as0TPPPKOffvpJmzdvNj5nJpNRJpMpOM6ePQCYWXDsJycnVV9ff+tn3/c1NTU159jx8XHF43EtW7ZMiURC8Xh8znG9vb3q6ekpOP73//wvC50uADjprrGPx+OanJyc83cXLlyY94Wefvpp/fjjj3r44Yd17do17dy5U7W1tXrppZcKxra1tc35h+DfLv7XvK8HAPh/d439yZMn7/j7uro6Xb9+XaFQSJKUSqX04osvFoxbsWLFrX83NjYqEonol19+mTP2wWBQwWCw4HjuP3+723QBAHMILPQEmzZtUn9/v6T/26a5ePGiXn311YJxMzMzt75p6saNGzp//ryefPLJe7pWLp8v+gAAFLfgPft33nlH7e3tamlpUSAQ0IEDB26t4o8cOaKamhq99dZbGh4e1vHjx7Vs2TJls1lt27ZNkUjknq7F3TgAYKYqb9EXu/7TqZGiv/vHrff2h2OxTBw8XO4poELdHPqh3FNABXv83NCinp930AKAAyyLPffZA4AJy2LPyh4ATFgVe75wHADMWBV7PggNAMxYFXvupwcAM3bFnm0cADBiVey5GwcAzFgWe1b2AGDCrtjzAi0AGLEq9uzZA4AZq2LPnj0AmLEs9qzsAcCEZbFnZQ8AJqyKPet6ADBjVexZ2QOAGctiz9oeAExYFfscK3sAMGJV7FnZA4AZu2LPO2gBwIhdsWdlDwBGrIo9e/YAYMau2PPlJQBgxKrYs40DAGasin2OF2gBwEhgoSc4deqUXn/9dT311FPq6+u749hvvvlGLS0tikQiOnDgwD3vwWdz+aIPAHBFd3e3NmzYoLVr1+ry5cvzes6CYx8Oh/XZZ58pFovdcdy1a9fU09Oj/v5+DQ8P648//tC33357T9fK5nJFHwDgiubmZh07dkz19fXzfs6Ct3GeeOIJSVIgcOe/G0NDQ4pEIgqFQpKk7du368SJE9q2bVvB2Ewmo0wmU3CcFTwASOvXr7/n5yzZnn0qlVJdXd2tn+vq6pRKpeYc29vbq56enoLj77//vvbu3btocwSA+9VdYx+PxzU5OTnn7y5cuCDP80o+qba2NsXj8YLjwWCw5NcCABfcNfYnT54syYV837/tj8bk5KR8359zbDAYJOwAUEILfoF2vqLRqEZGRpROp5XL5TQwMKDNmzcv1eUBwGkLjn0ymVRTU5MGBwd15MgRNTU16bfffpMkHTlyRMePH5ckNTY2avfu3dqxY4c2btyohoYGvfHGGwu9PAA459ChQ2pqatLU1JR27typLVu23PU5Vfk8n0FQShMHD5d7CqhQN4d+KPcUUMEePze0qOdfsm0cAED5EHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcACxBwAHEHsAcEBVPp/Pl3sSAIDFxcoeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQeABxA7AHAAcQei2JsbEytra2KRqNqbW3V+Ph4uaeECtHd3a0NGzZo7dq1unz5crmn4wxij0XR2dmpRCKhoaEhJRIJdXR0lHtKqBDNzc06duyY6uvryz0VpxB7lNzs7KxGR0cVi8UkSbFYTKOjo0qn02WeGSrB+vXr5ft+uafhHGKPkkulUqqtrZXneZIkz/NUU1OjVCpV5pkB7iL2AOAAYo+S831f09PTymazkqRsNquZmRn+6w6UEbFHyVVXVyscDiuZTEqSksmkwuGwQqFQmWcGuItvqsKi+P3339Xe3q5MJqNgMKju7m49+uij5Z4WKsChQ4c0PDysP//8U6tWrdLKlSt15syZck/rvkfsAcABbOMAgAOIPQA4gNgDgAOIPQA4gNgDgAOIPQA4gNgDgAP+F9OzVxiYW3omAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotter.plot_correlation_matrix(\n",
    "    series=\"returns\",\n",
    "    resample_rule=eval_config[\"resample_rule\"],\n",
    "    mode=eval_config[\"mode\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "749d6378",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.663766Z",
     "start_time": "2022-02-23T13:02:46.663738Z"
    }
   },
   "outputs": [],
   "source": [
    "plotter.plot_effective_correlation_rank(\n",
    "    series=\"returns\",\n",
    "    resample_rule=eval_config[\"resample_rule\"],\n",
    "    mode=eval_config[\"mode\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4a7c60d9",
   "metadata": {},
   "source": [
    "## Model correlation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c5e7659a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.666041Z",
     "start_time": "2022-02-23T13:02:46.666020Z"
    }
   },
   "outputs": [],
   "source": [
    "plotter.plot_correlation_matrix(\n",
    "    series=\"pnl\",\n",
    "    resample_rule=eval_config[\"resample_rule\"],\n",
    "    mode=eval_config[\"mode\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b7938396",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.668296Z",
     "start_time": "2022-02-23T13:02:46.668268Z"
    }
   },
   "outputs": [],
   "source": [
    "plotter.plot_effective_correlation_rank(\n",
    "    series=\"pnl\",\n",
    "    resample_rule=eval_config[\"resample_rule\"],\n",
    "    mode=eval_config[\"mode\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "215ed5cc",
   "metadata": {},
   "source": [
    "## Aggregate model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c23ae6ee",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.671098Z",
     "start_time": "2022-02-23T13:02:46.671082Z"
    }
   },
   "outputs": [],
   "source": [
    "pnl_srs, pos_srs, aggregate_stats = evaluator.aggregate_models(\n",
    "    keys=selected,\n",
    "    mode=eval_config[\"mode\"],\n",
    "    target_volatility=eval_config[\"target_volatility\"],\n",
    ")\n",
    "display(aggregate_stats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c61ac477",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.673070Z",
     "start_time": "2022-02-23T13:02:46.673053Z"
    }
   },
   "outputs": [],
   "source": [
    "plotter.plot_sharpe_ratio_panel(keys=selected, mode=eval_config[\"mode\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e6ab0618",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.675444Z",
     "start_time": "2022-02-23T13:02:46.675426Z"
    },
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "plotter.plot_rets_signal_analysis(\n",
    "    keys=selected,\n",
    "    resample_rule=eval_config[\"resample_rule\"],\n",
    "    mode=eval_config[\"mode\"],\n",
    "    target_volatility=eval_config[\"target_volatility\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "17854931",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.678242Z",
     "start_time": "2022-02-23T13:02:46.678215Z"
    }
   },
   "outputs": [],
   "source": [
    "plotter.plot_performance(\n",
    "    keys=selected,\n",
    "    resample_rule=eval_config[\"resample_rule\"],\n",
    "    mode=eval_config[\"mode\"],\n",
    "    target_volatility=eval_config[\"target_volatility\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "27a3a550",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.681130Z",
     "start_time": "2022-02-23T13:02:46.681090Z"
    }
   },
   "outputs": [],
   "source": [
    "plotter.plot_rets_and_vol(\n",
    "    keys=selected,\n",
    "    resample_rule=eval_config[\"resample_rule\"],\n",
    "    mode=eval_config[\"mode\"],\n",
    "    target_volatility=eval_config[\"target_volatility\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "628273de",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.682218Z",
     "start_time": "2022-02-23T13:02:46.682202Z"
    }
   },
   "outputs": [],
   "source": [
    "if False:\n",
    "    plotter.plot_positions(\n",
    "        keys=selected,\n",
    "        mode=eval_config[\"mode\"],\n",
    "        target_volatility=eval_config[\"target_volatility\"],\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "00e9e05f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-02-23T13:02:46.683099Z",
     "start_time": "2022-02-23T13:02:46.683084Z"
    }
   },
   "outputs": [],
   "source": [
    "if False:\n",
    "    # Plot the returns and prediction for one or more models.\n",
    "    model_key = selected[:1]\n",
    "    plotter.plot_returns_and_predictions(\n",
    "        keys=model_key,\n",
    "        resample_rule=eval_config[\"resample_rule\"],\n",
    "        mode=eval_config[\"mode\"],\n",
    "    )"
   ]
  }
 ],
 "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": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}