1_add-derived-features.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "a1c07eac-dca9-454f-8dea-2e9695017fa9",
   "metadata": {},
   "source": [
    "# Add Cycle-Level Features\n",
    "We use a few different types of faetures that summarize cycles for the machine learning models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "06d4dc4b-cae5-4af4-b247-e3fef85cc3e4",
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "from matplotlib import pyplot as plt\n",
    "from battdat.postprocess.integral import CapacityPerCycle, StateOfCharge\n",
    "from battdat.postprocess.timing import CycleTimesSummarizer\n",
    "from battdat.data import CellDataset\n",
    "from concurrent.futures import ProcessPoolExecutor\n",
    "from functools import partial\n",
    "from pathlib import Path\n",
    "from tqdm import tqdm\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f0070ab-db5c-4ac8-9a2c-5a3666f540a0",
   "metadata": {},
   "source": [
    "Configuration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a503d6ae-daf0-4d23-a1e5-9470b8a4cadb",
   "metadata": {},
   "outputs": [],
   "source": [
    "hdf5_path = './data/hdf5/'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "16fccd76-bb30-43e1-aaaf-bbf0fbf006d4",
   "metadata": {},
   "source": [
    "## Load in an example cell\n",
    "Start with one to show off how this works"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "944ca2d1-7d94-4cdf-aca4-7cbca3a8f003",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = CellDataset.from_hdf('./data/hdf5/refined/batch_B28B_cell_2.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6fc8a205-762c-4e9e-aedb-4936c41be363",
   "metadata": {},
   "source": [
    "## Compute the Features\n",
    "Each is described using a different class"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a54771a7-0303-48bf-b378-e9922977c3a2",
   "metadata": {},
   "source": [
    "### Capacity Per Cycle\n",
    "Compute the amount of charge pushed into the battery and the "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "b7f5fc2b-c434-4eb5-ab7d-d9af5a3e795d",
   "metadata": {},
   "outputs": [],
   "source": [
    "CapacityPerCycle().add_summaries(dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "c656a8c6-87b0-422e-8e0e-1feb9f3b3dcf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Charge Capacity (A-hr)')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 350x250 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(3.5, 2.5))\n",
    "\n",
    "ax.plot(dataset.cycle_stats['capacity_charge'])\n",
    "\n",
    "ax.set_xlabel('Cycle')\n",
    "ax.set_ylabel('Charge Capacity (A-hr)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "178c7d37-4811-4e6b-9249-dab587ac8707",
   "metadata": {},
   "outputs": [
    {
     "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>cycle_number</th>\n",
       "      <th>energy_charge</th>\n",
       "      <th>capacity_charge</th>\n",
       "      <th>energy_discharge</th>\n",
       "      <th>capacity_discharge</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>4.336126e-07</td>\n",
       "      <td>1.250629e-07</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1.460777e+00</td>\n",
       "      <td>3.913739e-01</td>\n",
       "      <td>1.545621</td>\n",
       "      <td>0.420994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1.562767e+00</td>\n",
       "      <td>4.219581e-01</td>\n",
       "      <td>1.539038</td>\n",
       "      <td>0.419249</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>1.555165e+00</td>\n",
       "      <td>4.194255e-01</td>\n",
       "      <td>1.517108</td>\n",
       "      <td>0.413389</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>1.534272e+00</td>\n",
       "      <td>4.134041e-01</td>\n",
       "      <td>1.513609</td>\n",
       "      <td>0.412468</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cycle_number  energy_charge  capacity_charge  energy_discharge  \\\n",
       "0             0   4.336126e-07     1.250629e-07          0.000000   \n",
       "1             1   1.460777e+00     3.913739e-01          1.545621   \n",
       "2             2   1.562767e+00     4.219581e-01          1.539038   \n",
       "3             3   1.555165e+00     4.194255e-01          1.517108   \n",
       "4             4   1.534272e+00     4.134041e-01          1.513609   \n",
       "\n",
       "   capacity_discharge  \n",
       "0            0.000000  \n",
       "1            0.420994  \n",
       "2            0.419249  \n",
       "3            0.413389  \n",
       "4            0.412468  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset.cycle_stats.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c3706f5-741a-41db-aaac-f584d776d5ce",
   "metadata": {},
   "source": [
    "## Run for each cell\n",
    "Load in, create cycle summaries, save back"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "2b806c37-58ec-4fe1-8298-da1e517865ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_computers = [CapacityPerCycle(), StateOfCharge(), CycleTimesSummarizer()]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c3828279-d32f-4200-868f-9f4a91b0383b",
   "metadata": {},
   "source": [
    "Make a function we can run in parallel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "065446c5-3fb2-4738-8714-c799629d16b8",
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_features(file: Path, computers: list):\n",
    "    \"\"\"Add features to a certain file\"\"\"\n",
    "    from warnings import catch_warnings, filterwarnings\n",
    "    dataset = CellDataset.from_hdf(str(file))\n",
    "    with catch_warnings():\n",
    "        filterwarnings('ignore')\n",
    "        for feature in feature_computers:\n",
    "            feature.compute_features(dataset)\n",
    "    dataset.to_hdf(str(file), complevel=9)\n",
    "    return file\n",
    "func = partial(add_features, computers=feature_computers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "bff61e7a-178d-4e74-be7c-074396f1dc75",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "602it [1:07:05,  6.69s/it]\n"
     ]
    }
   ],
   "source": [
    "for path in tqdm(Path(hdf5_path).rglob('*.h5')):\n",
    "    func(path)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "869421e8-207e-487a-8d3a-4c70788a0adc",
   "metadata": {},
   "source": [
    "## Show the improved HDF5 file\n",
    "The features are now part of HDF5 file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "43c10023-e3ca-42f7-b34e-c740ee964909",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = CellDataset.from_hdf('./data/hdf5/refined/batch_B28B_cell_2.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "75535afa-cf0e-486e-8529-290c5d6b090c",
   "metadata": {},
   "outputs": [
    {
     "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>cycle_number</th>\n",
       "      <th>energy_charge</th>\n",
       "      <th>capacity_charge</th>\n",
       "      <th>energy_discharge</th>\n",
       "      <th>capacity_discharge</th>\n",
       "      <th>cycle_start</th>\n",
       "      <th>cycle_duration</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>4.336126e-07</td>\n",
       "      <td>1.250629e-07</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>60.048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1.460777e+00</td>\n",
       "      <td>3.913739e-01</td>\n",
       "      <td>1.545621</td>\n",
       "      <td>0.420994</td>\n",
       "      <td>60.048</td>\n",
       "      <td>156219.450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1.562767e+00</td>\n",
       "      <td>4.219581e-01</td>\n",
       "      <td>1.539038</td>\n",
       "      <td>0.419249</td>\n",
       "      <td>156279.498</td>\n",
       "      <td>161755.470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>1.555165e+00</td>\n",
       "      <td>4.194255e-01</td>\n",
       "      <td>1.517108</td>\n",
       "      <td>0.413389</td>\n",
       "      <td>318034.968</td>\n",
       "      <td>80239.932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>1.534272e+00</td>\n",
       "      <td>4.134041e-01</td>\n",
       "      <td>1.513609</td>\n",
       "      <td>0.412468</td>\n",
       "      <td>398274.900</td>\n",
       "      <td>79615.722</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cycle_number  energy_charge  capacity_charge  energy_discharge  \\\n",
       "0             0   4.336126e-07     1.250629e-07          0.000000   \n",
       "1             1   1.460777e+00     3.913739e-01          1.545621   \n",
       "2             2   1.562767e+00     4.219581e-01          1.539038   \n",
       "3             3   1.555165e+00     4.194255e-01          1.517108   \n",
       "4             4   1.534272e+00     4.134041e-01          1.513609   \n",
       "\n",
       "   capacity_discharge  cycle_start  cycle_duration  \n",
       "0            0.000000        0.000          60.048  \n",
       "1            0.420994       60.048      156219.450  \n",
       "2            0.419249   156279.498      161755.470  \n",
       "3            0.413389   318034.968       80239.932  \n",
       "4            0.412468   398274.900       79615.722  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset.cycle_stats.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b460e0ad-4ca5-4045-8392-d4d033f542ad",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2548b245-d1d9-40d1-8569-fe6838ae001d",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.11.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}