exp1.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "5966eee5-f11d-4d37-b6a9-39618ada54c5",
   "metadata": {},
   "source": [
    "# Experiment 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c968559b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:03.543345Z",
     "iopub.status.busy": "2024-02-09T05:22:03.542803Z",
     "iopub.status.idle": "2024-02-09T05:22:04.039082Z",
     "shell.execute_reply": "2024-02-09T05:22:04.037956Z",
     "shell.execute_reply.started": "2024-02-09T05:22:03.543317Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import functools\n",
    "import pathlib\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import pickle\n",
    "from scipy import stats\n",
    "import seaborn as sns\n",
    "\n",
    "from videoannotator import config as cfg, experiments, data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "b3deafa3-cbd3-407e-a1b6-eac5d7151c0e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:04.041820Z",
     "iopub.status.busy": "2024-02-09T05:22:04.040603Z",
     "iopub.status.idle": "2024-02-09T05:22:04.046289Z",
     "shell.execute_reply": "2024-02-09T05:22:04.045327Z",
     "shell.execute_reply.started": "2024-02-09T05:22:04.041776Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "sns.set_context('notebook')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6208beed-69bb-450d-be7d-6923fac94cf8",
   "metadata": {},
   "source": [
    "## helpers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "f2d158c3-d8fd-49c2-b345-db6f12531592",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:14.662023Z",
     "iopub.status.busy": "2024-02-09T05:22:14.661411Z",
     "iopub.status.idle": "2024-02-09T05:22:14.697221Z",
     "shell.execute_reply": "2024-02-09T05:22:14.696577Z",
     "shell.execute_reply.started": "2024-02-09T05:22:14.661995Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "@dataclass(frozen=True)\n",
    "class Beat:\n",
    "    champ: experiments.Metric\n",
    "    chall: experiments.Metric\n",
    "        \n",
    "    @property\n",
    "    def better(self):\n",
    "        if self.champ is None:\n",
    "            return True\n",
    "        return self.chall.mean > self.champ.mean\n",
    "    \n",
    "    @property\n",
    "    def sig(self):\n",
    "        if self.better:\n",
    "            if self.champ is None: return True\n",
    "            return (self.chall.mean - self.chall.std) > (self.champ.mean + self.champ.std)\n",
    "        return False\n",
    "    \n",
    "    @property\n",
    "    def lift(self):\n",
    "        if self.champ and self.chall:\n",
    "            return self.chall.mean / self.champ.mean - 1\n",
    "        \n",
    "    @property\n",
    "    def gain(self):\n",
    "        if self.champ and self.chall:\n",
    "            return self.chall.mean - self.champ.mean\n",
    "    \n",
    "    def __repr__(self):\n",
    "        return f\"Beat(better={self.better}, sig={self.sig})\"\n",
    "\n",
    "def at(l, cnt):\n",
    "    d = res[l]\n",
    "    cmp = {\n",
    "        method: er\n",
    "        for (n, method, scoring), er in d['cmp'].items()\n",
    "        if n == cnt and scoring == 'average_precision'\n",
    "    }\n",
    "    va = [\n",
    "        x\n",
    "        for x in d['checkpoints']\n",
    "        if len(x.lds_train) <= (cnt + 10)  # allow for up to n + 10\n",
    "    ]\n",
    "    va_ap = max(va, key=lambda x: x.average_precision.mean).average_precision if len(va) > 0 else None\n",
    "    return dict(\n",
    "        label=l,\n",
    "        baseline=d['baseline_agg'],\n",
    "        zs=d['zero_shot'].average_precision,\n",
    "        va_cnt=len(va[-1].lds_train) if len(va) > 0 else None,\n",
    "        va=va_ap,\n",
    "        cmp_zs=cmp['zero-shot'].average_precision if 'zero-shot' in cmp else None,\n",
    "        cmp_random=cmp['random'].average_precision if 'random' in cmp else None,\n",
    "        cmp_comb=cmp['zero-shot-50-random-50'].average_precision if 'zero-shot-50-random-50' in cmp else None,\n",
    "    )\n",
    "\n",
    "def beat_cols(champ_, chall_):\n",
    "    return [\n",
    "        Beat(champ=champ, chall=chall)\n",
    "        for champ, chall in zip(champ_, chall_)\n",
    "    ]\n",
    "\n",
    "def analyze(cnt):\n",
    "    df = pd.DataFrame(map(lambda l: at(l, cnt=cnt), res.keys()))\n",
    "    n = len(df)\n",
    "    df = df[df.va_cnt.notnull()]\n",
    "    if len(df) < n:\n",
    "        print(f'lost {n - len(df)} / {n} for cnt= {cnt}')\n",
    "    return df.assign(\n",
    "        va_beats_bl=beat_cols(df.baseline, df.va),\n",
    "        va_beats_zs=beat_cols(df.zs, df.va),\n",
    "        va_beats_cmp_zs=beat_cols(df.cmp_zs, df.va),\n",
    "        va_beats_cmp_rand=beat_cols(df.cmp_random, df.va),\n",
    "        va_beats_cmp_comb=beat_cols(df.cmp_comb, df.va),\n",
    "        cnt=cnt,\n",
    "    )\n",
    "\n",
    "def agg(df, sig):\n",
    "    df = df.groupby('cnt').apply(\n",
    "        lambda rows: pd.Series(dict(\n",
    "            beats_baseline=rows.va_beats_bl.apply(lambda x: x.better).mean(),\n",
    "            beats_baseline_sig=rows.va_beats_bl.apply(lambda x: x.sig).mean(),\n",
    "            beats_zs=rows.va_beats_zs.apply(lambda x: x.better).mean(),\n",
    "            beats_zs_sig=rows.va_beats_zs.apply(lambda x: x.sig).mean(),\n",
    "            beats_cmp_rand=rows.va_beats_cmp_rand.apply(lambda x: x.better).mean(),\n",
    "            beats_cmp_rand_sig=rows.va_beats_cmp_rand.apply(lambda x: x.sig).mean(),\n",
    "            beats_cmp_zs=rows.va_beats_cmp_zs.apply(lambda x: x.better).mean(),\n",
    "            beats_cmp_zs_sig=rows.va_beats_cmp_zs.apply(lambda x: x.sig).mean(),\n",
    "            beats_cmp_comb=rows.va_beats_cmp_comb.apply(lambda x: x.better).mean(),\n",
    "            beats_cmp_comb_sig=rows.va_beats_cmp_comb.apply(lambda x: x.sig).mean(),\n",
    "        ))\n",
    "    )\n",
    "    cols = [c for c in df.columns if ((sig and '_sig' in c) or (not sig and '_sig' not in c))]\n",
    "    return df[cols]\n",
    "\n",
    "\n",
    "def _get_diff(df, c):\n",
    "    return [\n",
    "        x.chall.mean - x.champ.mean\n",
    "        for x in df[c]\n",
    "        if x.lift\n",
    "    ]\n",
    "\n",
    "def f(df, col=None):\n",
    "    rows = []\n",
    "\n",
    "    cols = [c for c in d.columns if 'va_beats_' in c]\n",
    "    cs = cols if col is None else [col]\n",
    "    for c in cs:\n",
    "        gain = _get_diff(df, c)\n",
    "        if len(gain) == 0:\n",
    "            x = 'nan'\n",
    "            continue\n",
    "        x = np.percentile(gain, q=50)\n",
    "        rows.append(dict(c=c, val=x))\n",
    "\n",
    "    if col is None:\n",
    "        return pd.DataFrame(rows).set_index('c')\n",
    "    else:\n",
    "        return pd.Series([x['val'] for x in rows], index=[x['c'] for x in rows])\n",
    "    \n",
    "def coverage(df):\n",
    "    cs = ['va_beats_cmp_zs', 'va_beats_cmp_rand', 'va_beats_cmp_comb']\n",
    "    return pd.Series([\n",
    "        sum(\n",
    "            x.champ is not None\n",
    "            for x in df[c]\n",
    "        ) / len(cfg.LABELS)\n",
    "        for c in cs\n",
    "    ], index=cs)\n",
    "\n",
    "\n",
    "def lift(df, q, col=None):\n",
    "    cs = [c for c in cols if 'bl' not in c] if col is None else [col]\n",
    "    return pd.Series([\n",
    "        np.percentile(_get_diff(df, c), q=q)\n",
    "        for c in cs\n",
    "    ], index=cs)\n",
    "\n",
    "def plot(res, label):\n",
    "    try:\n",
    "        pd.DataFrame(\n",
    "            dict(\n",
    "                n=n,\n",
    "                method=method,\n",
    "                r=r.average_precision.mean,\n",
    "            )\n",
    "            for (n, method, scoring), r in res[label]['cmp'].items()\n",
    "            if scoring == 'average_precision'\n",
    "        ).pivot_table(index='n', columns='method', values='r').plot()\n",
    "    except Exception as e:\n",
    "        print(f'failed {label} error: {e}')\n",
    "    pd.DataFrame(\n",
    "        dict(n=len(er.lds_train), ap=er.average_precision.mean)\n",
    "        for er in res[label]['checkpoints']\n",
    "    ).set_index('n').ap.plot()\n",
    "    plt.hlines(\n",
    "        res[label]['baseline_agg'].mean,\n",
    "        xmin=0,\n",
    "        xmax=1000,\n",
    "        colors='r',\n",
    "        linestyles='dotted',\n",
    "    )\n",
    "    plt.hlines(\n",
    "        res[label]['zero_shot'].average_precision.mean,\n",
    "        xmin=0,\n",
    "        xmax=1000,\n",
    "        colors='k',\n",
    "        linestyles='dotted',\n",
    "    )\n",
    "    plt.ylim([0, 1])\n",
    "    plt.title(label)\n",
    "    plt.show()\n",
    "    \n",
    "@functools.lru_cache()\n",
    "def get_cmp_data_scores(l, which):\n",
    "    cmp = data.CompDataset(l, frozenset())\n",
    "    fn = getattr(cmp, f'get_ordered_dataset_{which}')\n",
    "    od = fn(keys_to_remove=frozenset())\n",
    "    return {n: od.get_n(n).lds.data_score for n in ns}\n",
    "\n",
    "@functools.lru_cache()\n",
    "def get_cmp_comb_data_scores(l):\n",
    "    cmp = data.CompDataset(l, frozenset())\n",
    "    return {\n",
    "        n: (cmp.get_ordered_dataset_random().get_n(n // 2) + cmp.get_ordered_dataset_zero_shot().get_n(n // 2)).lds.data_score\n",
    "        for n in ns\n",
    "    }\n",
    "\n",
    "@functools.lru_cache()\n",
    "def get_va_data_scores(l):\n",
    "    ldss = data.get_labeled_dataset_checkpoints(l)\n",
    "    vals = dict()\n",
    "    d = 10\n",
    "    for lds in ldss:\n",
    "        if len(lds) >= d:\n",
    "            vals[(len(lds) // d) * d] = lds.data_score\n",
    "    return [\n",
    "        dict(\n",
    "            label=l,\n",
    "            n=k,\n",
    "            method='Video Annotator (VA)',\n",
    "            data_score=v,\n",
    "        )\n",
    "        for k, v in vals.items()\n",
    "    ]\n",
    "\n",
    "group_lookup = {\n",
    "    v: k\n",
    "    for k, vs in cfg.LABEL_GROUPS.items()\n",
    "    for v in vs\n",
    "}"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7564f811-8148-4942-a26f-1fc34fc679e6",
   "metadata": {},
   "source": [
    "## run experiment 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "48b4ce7a-1411-4b48-b85d-a6bab2249065",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:19.266696Z",
     "iopub.status.busy": "2024-02-09T05:22:19.266267Z",
     "iopub.status.idle": "2024-02-09T05:22:19.695469Z",
     "shell.execute_reply": "2024-02-09T05:22:19.694457Z",
     "shell.execute_reply.started": "2024-02-09T05:22:19.266668Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# this takes ~20 mins on a machine with 64 CPUs and 64GB RAM\n",
    "path_results_pkl = 'exp1_results.pkl'\n",
    "if not pathlib.Path(path_results_pkl).exists():\n",
    "    exp = experiments.Experiment()\n",
    "    res = exp.experiment_active_learning_batch(labels=cfg.LABELS)\n",
    "    pickle.dump(res, open(path_results_pkl, 'wb'))\n",
    "else:\n",
    "    res = pickle.load(open(path_results_pkl, 'rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c72aa7b2-5aaf-447a-8cff-0a9444856599",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:21.425445Z",
     "iopub.status.busy": "2024-02-09T05:22:21.425032Z",
     "iopub.status.idle": "2024-02-09T05:22:21.613075Z",
     "shell.execute_reply": "2024-02-09T05:22:21.612433Z",
     "shell.execute_reply.started": "2024-02-09T05:22:21.425416Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING 2024-02-09 05:22:21,498 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,498 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,499 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,504 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,504 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,505 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,505 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,506 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,506 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,569 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,569 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,570 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,570 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,571 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,571 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,572 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,573 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,574 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,574 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,575 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,575 experiments.py(32) Metric has NaN values: (0.2, nan, 0.3333333333333333, 0.2857142857142857, 0.3333333333333333)\n",
      "WARNING 2024-02-09 05:22:21,576 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,576 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:21,577 experiments.py(32) Metric has NaN values: (0.2, nan, 0.3333333333333333, 0.2857142857142857, 0.3333333333333333)\n",
      "WARNING 2024-02-09 05:22:21,577 experiments.py(32) Metric has NaN values: (0.2, nan, 0.3333333333333333, 0.2857142857142857, 0.3333333333333333)\n",
      "WARNING 2024-02-09 05:22:21,578 experiments.py(32) Metric has NaN values: (0.2, nan, 0.3333333333333333, 0.2857142857142857, 0.3333333333333333)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lost 23 / 56 for cnt= 25\n",
      "lost 7 / 56 for cnt= 50\n",
      "lost 4 / 56 for cnt= 100\n"
     ]
    }
   ],
   "source": [
    "# organize the data for analysis\n",
    "df = pd.concat([analyze(n) for n in (25, 50, 100, 500, 1000)])\n",
    "d = df.assign(group=df.label.map(group_lookup))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "200de408-d2f3-45e1-a9f7-e57b25863f7b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T02:12:48.296946Z",
     "iopub.status.busy": "2024-02-09T02:12:48.296440Z",
     "iopub.status.idle": "2024-02-09T02:12:48.300481Z",
     "shell.execute_reply": "2024-02-09T02:12:48.299879Z",
     "shell.execute_reply.started": "2024-02-09T02:12:48.296918Z"
    },
    "tags": []
   },
   "source": [
    "## Fig 7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "2f037e99",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:24.338170Z",
     "iopub.status.busy": "2024-02-09T05:22:24.337652Z",
     "iopub.status.idle": "2024-02-09T05:22:24.790900Z",
     "shell.execute_reply": "2024-02-09T05:22:24.790262Z",
     "shell.execute_reply.started": "2024-02-09T05:22:24.338141Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cov = d.groupby('cnt').apply(coverage) * 100\n",
    "cov.rename(columns=dict(\n",
    "    va_beats_cmp_zs='Zero Shot Classifier (ZSC)',\n",
    "    va_beats_cmp_rand='Randomly-selected Clip Classifier (RCC)',\n",
    "    va_beats_cmp_comb='Combined Classifier (CC)',\n",
    ")).plot()\n",
    "plt.xlabel('Sample size ($n$)')\n",
    "_ = plt.ylabel('Percentage of labels with enough data')\n",
    "_ = plt.ylim([0, 100])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "132d949e-884b-428a-acb3-d5189faf350b",
   "metadata": {},
   "source": [
    "## Fig 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "b69fff90",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:26.441912Z",
     "iopub.status.busy": "2024-02-09T05:22:26.441153Z",
     "iopub.status.idle": "2024-02-09T05:22:26.616036Z",
     "shell.execute_reply": "2024-02-09T05:22:26.615420Z",
     "shell.execute_reply.started": "2024-02-09T05:22:26.441883Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "label = 'establishing-shots'\n",
    "plot(res, label)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "38568d1c-ca83-4a18-a1fb-60fc11626306",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T02:11:30.360649Z",
     "iopub.status.busy": "2024-02-09T02:11:30.360132Z",
     "iopub.status.idle": "2024-02-09T02:11:30.364332Z",
     "shell.execute_reply": "2024-02-09T02:11:30.363738Z",
     "shell.execute_reply.started": "2024-02-09T02:11:30.360620Z"
    },
    "tags": []
   },
   "source": [
    "## Fig 9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "08c30a2b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:27.999138Z",
     "iopub.status.busy": "2024-02-09T05:22:27.997926Z",
     "iopub.status.idle": "2024-02-09T05:22:29.115929Z",
     "shell.execute_reply": "2024-02-09T05:22:29.115270Z",
     "shell.execute_reply.started": "2024-02-09T05:22:27.999068Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING 2024-02-09 05:22:28,610 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:28,670 experiments.py(32) Metric has NaN values: (0.1, nan, 0.1111111111111111, 0.15384615384615385, 0.2222222222222222)\n",
      "WARNING 2024-02-09 05:22:28,675 experiments.py(32) Metric has NaN values: (0.5, nan, 1.0, 0.6666666666666666, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:28,685 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:28,685 experiments.py(32) Metric has NaN values: (0.5, nan, 1.0, 0.6666666666666666, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:28,820 experiments.py(32) Metric has NaN values: (0.1, nan, 0.1111111111111111, 0.15384615384615385, 0.2222222222222222)\n",
      "WARNING 2024-02-09 05:22:28,826 experiments.py(32) Metric has NaN values: (0.5, nan, 1.0, 0.6666666666666666, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:28,835 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:28,835 experiments.py(32) Metric has NaN values: (0.5, nan, 1.0, 0.6666666666666666, 0.6666666666666666)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "q = (df.groupby('cnt').apply(lambda xs: pd.Series({\n",
    "    c: sum(\n",
    "        (x.lift and x.lift >= 0) or (x.chall is not None and x.champ is None)\n",
    "        for x in xs[c]\n",
    "    ) / sum(\n",
    "        x.chall is not None\n",
    "        for x in xs[c]\n",
    "    )\n",
    "    for c in df.columns if '_beats_' in c\n",
    "})) * 100)\n",
    "q.rename(columns={\n",
    "    'va_beats_bl': 'Baseline (B)',\n",
    "    'va_beats_zs': 'Zero Shot (ZS)',\n",
    "    'va_beats_cmp_zs': 'Zero Shot Classifier (ZSC)',\n",
    "    'va_beats_cmp_rand': 'Randomly-selected Clip Classifier (RCC)',\n",
    "    'va_beats_cmp_comb': 'Combined Classifier (CC)',\n",
    "}).plot()\n",
    "plt.ylim([50, 100])\n",
    "plt.xlabel('Sample size ($n$)')\n",
    "_ = plt.ylabel('Percentage of labels where VA beats the method')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "0100bd0e-2522-402c-b4ce-e282b3f29bd5",
   "metadata": {},
   "source": [
    "## Table 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "2048b700",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:31.632076Z",
     "iopub.status.busy": "2024-02-09T05:22:31.631633Z",
     "iopub.status.idle": "2024-02-09T05:22:31.772826Z",
     "shell.execute_reply": "2024-02-09T05:22:31.771909Z",
     "shell.execute_reply.started": "2024-02-09T05:22:31.632039Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING 2024-02-09 05:22:31,718 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:31,727 experiments.py(32) Metric has NaN values: (0.5, nan, 1.0, 0.6666666666666666, 0.6666666666666666)\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>cnt</th>\n",
       "      <th>25</th>\n",
       "      <th>50</th>\n",
       "      <th>100</th>\n",
       "      <th>1000</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>group</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>emotions</th>\n",
       "      <td>-6.0</td>\n",
       "      <td>-5.1</td>\n",
       "      <td>-4.2</td>\n",
       "      <td>3.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>events_actions</th>\n",
       "      <td>6.9</td>\n",
       "      <td>5.5</td>\n",
       "      <td>5.7</td>\n",
       "      <td>3.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>focus</th>\n",
       "      <td>1.2</td>\n",
       "      <td>-1.2</td>\n",
       "      <td>-0.4</td>\n",
       "      <td>1.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>genres</th>\n",
       "      <td>6.6</td>\n",
       "      <td>5.3</td>\n",
       "      <td>3.9</td>\n",
       "      <td>1.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>motion</th>\n",
       "      <td>31.5</td>\n",
       "      <td>1.2</td>\n",
       "      <td>5.6</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensitivities</th>\n",
       "      <td>0.5</td>\n",
       "      <td>-0.3</td>\n",
       "      <td>0.8</td>\n",
       "      <td>3.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>shot_types</th>\n",
       "      <td>1.2</td>\n",
       "      <td>0.6</td>\n",
       "      <td>0.5</td>\n",
       "      <td>2.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time_location</th>\n",
       "      <td>4.6</td>\n",
       "      <td>-8.5</td>\n",
       "      <td>1.3</td>\n",
       "      <td>1.4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "cnt             25    50    100   1000\n",
       "group                                 \n",
       "emotions        -6.0  -5.1  -4.2   3.7\n",
       "events_actions   6.9   5.5   5.7   3.8\n",
       "focus            1.2  -1.2  -0.4   1.8\n",
       "genres           6.6   5.3   3.9   1.8\n",
       "motion          31.5   1.2   5.6   3.0\n",
       "sensitivities    0.5  -0.3   0.8   3.4\n",
       "shot_types       1.2   0.6   0.5   2.3\n",
       "time_location    4.6  -8.5   1.3   1.4"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(d[d.cnt != 500].groupby(['cnt', 'group']).apply(lambda x: f(x, col='va_beats_cmp_comb')).unstack().T * 100).round(1).droplevel(0) #.to_latex()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "71f211f1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:37.361357Z",
     "iopub.status.busy": "2024-02-09T05:22:37.360924Z",
     "iopub.status.idle": "2024-02-09T05:22:37.396071Z",
     "shell.execute_reply": "2024-02-09T05:22:37.395355Z",
     "shell.execute_reply.started": "2024-02-09T05:22:37.361329Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING 2024-02-09 05:22:37,380 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:37,381 experiments.py(32) Metric has NaN values: (0.5, nan, 1.0, 0.6666666666666666, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:22:37,387 experiments.py(32) Metric has NaN values: (0.5, nan, 1.0, 0.6666666666666666, 0.6666666666666666)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "                   cnt \n",
       "va_beats_cmp_comb  25      1.2\n",
       "                   50      0.4\n",
       "                   100     1.5\n",
       "                   500     1.7\n",
       "                   1000    2.9\n",
       "dtype: float64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# last row in table 2\n",
    "(d.groupby(['cnt']).apply(lambda x: f(x, col='va_beats_cmp_comb')).unstack().T * 100).round(1)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "a96ed85c-4ff2-48e6-93ef-8b4737358676",
   "metadata": {},
   "source": [
    "## Fig 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "82f0c9fe-b923-490b-b0fe-46ef2e70f0be",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:44.829312Z",
     "iopub.status.busy": "2024-02-09T05:22:44.828905Z",
     "iopub.status.idle": "2024-02-09T05:22:57.095124Z",
     "shell.execute_reply": "2024-02-09T05:22:57.094210Z",
     "shell.execute_reply.started": "2024-02-09T05:22:44.829285Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING 2024-02-09 05:22:47,667 data.py(64) LabeledDataset(label=shutter-shot, cnt=25, pos_cnt=0, neg_cnt=25, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:47,672 data.py(64) LabeledDataset(label=shutter-shot, cnt=50, pos_cnt=0, neg_cnt=50, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:47,676 data.py(64) LabeledDataset(label=shutter-shot, cnt=100, pos_cnt=0, neg_cnt=100, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:47,680 data.py(64) LabeledDataset(label=shutter-shot, cnt=500, pos_cnt=0, neg_cnt=500, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:47,684 data.py(64) LabeledDataset(label=shutter-shot, cnt=1000, pos_cnt=0, neg_cnt=1000, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:47,761 data.py(64) LabeledDataset(label=timelapse, cnt=25, pos_cnt=0, neg_cnt=25, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:47,764 data.py(64) LabeledDataset(label=timelapse, cnt=50, pos_cnt=0, neg_cnt=50, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:47,768 data.py(64) LabeledDataset(label=timelapse, cnt=100, pos_cnt=0, neg_cnt=100, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:47,772 data.py(64) LabeledDataset(label=timelapse, cnt=500, pos_cnt=0, neg_cnt=500, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:47,776 data.py(64) LabeledDataset(label=timelapse, cnt=1000, pos_cnt=0, neg_cnt=1000, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,465 data.py(64) LabeledDataset(label=alcohol, cnt=25, pos_cnt=0, neg_cnt=25, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,472 data.py(64) LabeledDataset(label=alcohol, cnt=50, pos_cnt=0, neg_cnt=50, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,477 data.py(64) LabeledDataset(label=alcohol, cnt=100, pos_cnt=0, neg_cnt=100, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,483 data.py(64) LabeledDataset(label=alcohol, cnt=500, pos_cnt=0, neg_cnt=500, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,487 data.py(64) LabeledDataset(label=alcohol, cnt=1000, pos_cnt=0, neg_cnt=1000, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,597 data.py(64) LabeledDataset(label=jump-scare, cnt=25, pos_cnt=0, neg_cnt=25, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,602 data.py(64) LabeledDataset(label=jump-scare, cnt=50, pos_cnt=0, neg_cnt=50, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,607 data.py(64) LabeledDataset(label=jump-scare, cnt=100, pos_cnt=0, neg_cnt=100, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,613 data.py(64) LabeledDataset(label=jump-scare, cnt=500, pos_cnt=0, neg_cnt=500, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,619 data.py(64) LabeledDataset(label=jump-scare, cnt=1000, pos_cnt=0, neg_cnt=1000, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,824 data.py(64) LabeledDataset(label=shutter-shot, cnt=25, pos_cnt=0, neg_cnt=25, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,828 data.py(64) LabeledDataset(label=shutter-shot, cnt=50, pos_cnt=0, neg_cnt=50, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,831 data.py(64) LabeledDataset(label=shutter-shot, cnt=100, pos_cnt=0, neg_cnt=100, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,835 data.py(64) LabeledDataset(label=shutter-shot, cnt=500, pos_cnt=0, neg_cnt=500, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:48,839 data.py(64) LabeledDataset(label=shutter-shot, cnt=1000, pos_cnt=0, neg_cnt=1000, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:50,227 data.py(64) LabeledDataset(label=shutter-shot, cnt=24, pos_cnt=0, neg_cnt=24, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:50,238 data.py(64) LabeledDataset(label=shutter-shot, cnt=50, pos_cnt=0, neg_cnt=50, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:50,247 data.py(64) LabeledDataset(label=shutter-shot, cnt=100, pos_cnt=0, neg_cnt=100, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:50,257 data.py(64) LabeledDataset(label=shutter-shot, cnt=500, pos_cnt=0, neg_cnt=500, pos_rate=0.00) has no positive annotations.\n",
      "WARNING 2024-02-09 05:22:50,273 data.py(64) LabeledDataset(label=shutter-shot, cnt=1000, pos_cnt=0, neg_cnt=1000, pos_rate=0.00) has no positive annotations.\n"
     ]
    }
   ],
   "source": [
    "ns = (25, 50, 100, 500, 1_000)\n",
    "df_ds = pd.concat([\n",
    "    pd.DataFrame(\n",
    "        dict(\n",
    "            n=n,\n",
    "            method='Upper bound',\n",
    "            data_score=min(n / 200, 1),\n",
    "        )\n",
    "        for n in ns\n",
    "    ),\n",
    "    pd.DataFrame(\n",
    "        dict(n=n, method='Zero Shot Classifier (ZSC)', data_score=get_cmp_data_scores(l=l, which='zero_shot')[n])\n",
    "        for n in ns\n",
    "        for l in cfg.LABELS\n",
    "    ),\n",
    "    pd.DataFrame(\n",
    "        dict(n=n, method='Randomly-selected Clip Classifier (RCC)', data_score=get_cmp_data_scores(l=l, which='random')[n])\n",
    "        for n in ns\n",
    "        for l in cfg.LABELS\n",
    "    ),\n",
    "    pd.DataFrame(\n",
    "        dict(n=n, method='Combined Classifier (CC)', data_score=get_cmp_comb_data_scores(l)[n])\n",
    "        for n in ns\n",
    "        for l in cfg.LABELS\n",
    "    ),\n",
    "    pd.DataFrame(\n",
    "        row\n",
    "        for l in cfg.LABELS\n",
    "        for row in get_va_data_scores(l)\n",
    "    )\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "4da7467c-5cd8-4bdc-a73e-07f486aae0dd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:22:57.097094Z",
     "iopub.status.busy": "2024-02-09T05:22:57.096722Z",
     "iopub.status.idle": "2024-02-09T05:23:00.588308Z",
     "shell.execute_reply": "2024-02-09T05:23:00.587667Z",
     "shell.execute_reply.started": "2024-02-09T05:22:57.097067Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.lineplot(data=df_ds, x='n', y='data_score', hue='method')\n",
    "plt.ylim([0, 1.05])\n",
    "plt.ylabel('Data score')\n",
    "_ = plt.xlabel('Sample size ($n$)')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "28a21250-5c94-4d35-97aa-063c37e65263",
   "metadata": {},
   "source": [
    "## store values needed for experiment 2"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "dde0e28e-76cf-4e4e-af18-dcbd61e12d2e",
   "metadata": {},
   "source": [
    "### store last and best values per label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7b5a029b-7359-472b-9726-75d28ddf5e83",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:23:03.973610Z",
     "iopub.status.busy": "2024-02-09T05:23:03.973187Z",
     "iopub.status.idle": "2024-02-09T05:23:04.084827Z",
     "shell.execute_reply": "2024-02-09T05:23:04.084202Z",
     "shell.execute_reply.started": "2024-02-09T05:23:03.973583Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING 2024-02-09 05:23:03,979 experiments.py(32) Metric has NaN values: (0.2, nan, 0.3333333333333333, 0.2857142857142857, 0.3333333333333333)\n",
      "WARNING 2024-02-09 05:23:03,980 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,981 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,981 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,982 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,982 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,983 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,983 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,984 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,984 experiments.py(32) Metric has NaN values: (0.25, nan, 0.5, 0.3333333333333333, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,985 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,985 experiments.py(32) Metric has NaN values: (0.2, nan, 0.3333333333333333, 0.2857142857142857, 0.3333333333333333)\n",
      "WARNING 2024-02-09 05:23:03,986 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,986 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:03,987 experiments.py(32) Metric has NaN values: (0.2, nan, 0.3333333333333333, 0.2857142857142857, 0.3333333333333333)\n",
      "WARNING 2024-02-09 05:23:03,987 experiments.py(32) Metric has NaN values: (0.2, nan, 0.3333333333333333, 0.2857142857142857, 0.3333333333333333)\n"
     ]
    }
   ],
   "source": [
    "ap_va_best = {\n",
    "    l: dict(\n",
    "        last=res[l]['checkpoints'][-1].average_precision.mean,\n",
    "        best=max(\n",
    "            res[l]['checkpoints'],\n",
    "            key=lambda x: (\n",
    "                x.average_precision.mean\n",
    "                if x.average_precision.mean is not None\n",
    "                else -1\n",
    "            ),\n",
    "        ).average_precision.mean\n",
    "    )\n",
    "    for l in cfg.LABELS\n",
    "}\n",
    "# this object is used for experiment 2\n",
    "pickle.dump(ap_va_best, open('ap_va_best.pkl', 'wb'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "834fb5f8-063a-47c4-8881-5bc98309d4d7",
   "metadata": {},
   "source": [
    "### per-label AP gain at n=1,000\n",
    "for labels with max(n) < 1000, we just take the last value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "4c2d7ef9-6754-498a-849e-90747dfbd6a9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-02-09T05:23:05.766157Z",
     "iopub.status.busy": "2024-02-09T05:23:05.765707Z",
     "iopub.status.idle": "2024-02-09T05:23:05.811234Z",
     "shell.execute_reply": "2024-02-09T05:23:05.810162Z",
     "shell.execute_reply.started": "2024-02-09T05:23:05.766120Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING 2024-02-09 05:23:05,783 experiments.py(32) Metric has NaN values: (0.3333333333333333, nan, 0.5, 0.4, 0.6666666666666666)\n",
      "WARNING 2024-02-09 05:23:05,788 experiments.py(32) Metric has NaN values: (0.5, nan, 1.0, 0.6666666666666666, 0.6666666666666666)\n"
     ]
    }
   ],
   "source": [
    "d1000 = d[d.cnt == 1_000]\n",
    "q = d1000[['label']].assign(gain=d[d.cnt == 1000]['va_beats_cmp_comb'].apply(lambda x: x.gain)).set_index('label').gain.to_dict()\n",
    "pickle.dump(q, open('exp1_gain.pkl', 'wb'))"
   ]
  }
 ],
 "metadata": {
  "hide_input": false,
  "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": "3.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}