new_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "75924433",
   "metadata": {},
   "source": [
    "### Import"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "ab38ff56",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.container import BarContainer\n",
    "import seaborn as sns\n",
    "from scipy import stats\n",
    "from collections import defaultdict\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from os import listdir\n",
    "import json\n",
    "\n",
    "from utils.plotting import *"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "796a25e6",
   "metadata": {},
   "source": [
    "# Read data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "c47c71bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_data(folder):\n",
    "    # Initialize list of dataframes that we will concatenate.\n",
    "    all_data = []\n",
    "    \n",
    "    # Get list of all JSON results files in the folder.\n",
    "    files_to_read = [f for f in listdir(folder) if f.endswith(\".json\")]\n",
    "\n",
    "    # Read each file into a single dataframe.\n",
    "    for f in files_to_read:\n",
    "        with open(folder + \"/\" + f, \"r\") as fp:\n",
    "            data = json.load(fp)\n",
    "        results = pd.DataFrame(data[\"results\"])\n",
    "        for k, v in data[\"meta\"].items():\n",
    "            if not isinstance(v, list):\n",
    "                results[k] = v\n",
    "        results[\"corpus\"] = f.split(\"_\")[0]\n",
    "        all_data.append(results)\n",
    "    df = pd.concat(all_data)\n",
    "    \n",
    "    # Rename variables for better-looking figures.\n",
    "    df[\"model\"] = df.model.apply(pretty_model)\n",
    "    df[\"eval_type\"] = df.eval_type.apply(pretty_evaltype)\n",
    "\n",
    "    # Update model_name to include revision and quantization when applicable\n",
    "    df['model_with_revision'] = df.apply(\n",
    "        lambda row: f\"{row['model']}\" + \n",
    "                    (f\"-{row['quantization']}\" if row['quantization'] in [\"4bit\", \"8bit\"] else \"\") + \n",
    "                    (f\"\\n{row['revision']}\" if \"Pythia\" in row[\"model\"] or \"OLMo\" in row[\"model\"] else \"\") \n",
    "                    if \"Pythia\" in row[\"model\"] or \"OLMo\" in row[\"model\"] else row[\"model\"],\n",
    "        axis=1\n",
    "    )\n",
    "\n",
    "    return df\n",
    "\n",
    "# Read results for all tasks.\n",
    "RESULT_DIR = \"results\"\n",
    "TASKS = [\n",
    "    \"exp1_word-prediction\",\n",
    "    \"exp2_word-comparison\",\n",
    "    \"exp3a_sentence-judgment\",\n",
    "    \"exp3b_sentence-comparison\"\n",
    "]\n",
    "RESULTS = {\n",
    "    task.split(\"_\")[0]: read_data(f\"{RESULT_DIR}/{task}\")\n",
    "    for task in TASKS\n",
    "}\n",
    "\n",
    "# Copy direct data from Experiment 3a to 3b.\n",
    "_3a = RESULTS[\"exp3a\"].copy()\n",
    "_3a_direct = _3a[_3a.eval_type==\"Direct\"]\n",
    "_3b = RESULTS[\"exp3b\"].copy()\n",
    "_3b = pd.concat([_3a_direct, _3b])\n",
    "RESULTS[\"exp3b\"] = _3b"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "55a4b351-0e6a-4490-9157-bffc1a4c0eb9",
   "metadata": {},
   "source": [
    "# Figure 2: Task performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "5adad4e3-bb36-451f-a7ed-5d3791385889",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_accuracy_metric(df, task):\n",
    "    if task == \"exp1\":\n",
    "        pass\n",
    "    \n",
    "    elif task == \"exp2\":\n",
    "        df[\"model_prefers_good_continuation\"] = \\\n",
    "            (df[\"logprob_of_good_continuation\"] > df[\"logprob_of_bad_continuation\"])\n",
    "    \n",
    "    elif task == \"exp3b\":\n",
    "        def _success_criterion_sentence(row):\n",
    "            if row.eval_type == \"Direct\":\n",
    "                return (row.logprob_of_good_sentence > row.logprob_of_bad_sentence)\n",
    "            else:\n",
    "                return (row.logprob_of_good_continuation > row.logprob_of_bad_continuation)\n",
    "        df[\"model_prefers_good_sentence\"] = df.apply(_success_criterion_sentence, axis=1)\n",
    "    \n",
    "    else:\n",
    "        # Accuracy for \"direct\"\n",
    "        df[\"model_prefers_good_sentence\"] = df.apply(\n",
    "            lambda row: row.logprob_of_good_sentence > row.logprob_of_bad_sentence if row.eval_type == \"Direct\" else None,\n",
    "            axis=1\n",
    "        )\n",
    "        # True positive\n",
    "        df[\"model_prefers_yes_good\"] = df.apply(\n",
    "            lambda row: row.logprob_of_yes_good_sentence > row.logprob_of_no_good_sentence if row.eval_type != \"Direct\" else None,\n",
    "            axis=1\n",
    "        )\n",
    "        # True negative\n",
    "        df[\"model_prefers_no_bad\"] = df.apply(\n",
    "            lambda row: row.logprob_of_no_bad_sentence > row.logprob_of_yes_bad_sentence if row.eval_type != \"Direct\" else None,\n",
    "            axis=1\n",
    "        )\n",
    "        # False positive\n",
    "        df[\"model_prefers_yes_bad\"] = df.apply(\n",
    "            lambda row: row.logprob_of_yes_bad_sentence > row.logprob_of_no_bad_sentence if row.eval_type != \"Direct\" else None,\n",
    "            axis=1\n",
    "        )\n",
    "        # False negative\n",
    "        df[\"model_prefers_no_good\"] = df.apply(\n",
    "            lambda row: row.logprob_of_no_good_sentence > row.logprob_of_yes_good_sentence if row.eval_type != \"Direct\" else None,\n",
    "            axis=1\n",
    "        )\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "69106020",
   "metadata": {},
   "outputs": [],
   "source": [
    "for task, df in RESULTS.items():\n",
    "    RESULTS[task] = compute_accuracy_metric(df, task)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cbf9e658-accc-4ccd-be6c-4a2b61600b6f",
   "metadata": {},
   "source": [
    "## Experiment 1: Word prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6e660c46",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "d:\\Uni\\metalinguistic-prompting\\utils\\plotting.py:165: UserWarning: set_ticklabels() should only be used with a fixed number of ticks, i.e. after set_ticks() or using a FixedLocator.\n",
      "  ax.set_xticklabels(labels, rotation=25, ha=\"right\")\n",
      "d:\\Uni\\metalinguistic-prompting\\utils\\plotting.py:165: UserWarning: set_ticklabels() should only be used with a fixed number of ticks, i.e. after set_ticks() or using a FixedLocator.\n",
      "  ax.set_xticklabels(labels, rotation=25, ha=\"right\")\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "_df = RESULTS[\"exp1\"]\n",
    "metric = \"logprob_of_gold_continuation\"\n",
    "metric_label = \"Logprob of final word\"\n",
    "kwargs = dict(\n",
    "    x=\"model_with_revision\",\n",
    "    hue=\"eval_type\",\n",
    "    xlabel=\"\",\n",
    "    hue_order=EVAL_TYPES,\n",
    "    palette=EVAL_TYPE_PAL\n",
    ")\n",
    "\n",
    "########################################\n",
    "# Average over datasets\n",
    "########################################\n",
    "\n",
    "# ax = plot_metric(_df, metric, ylabel=metric_label, **kwargs)\n",
    "# ax.get_legend().remove()\n",
    "# plt.gcf().set_size_inches(*BARPLOT_SIZE)\n",
    "# # render(\"exp1_task_performance.pdf\")\n",
    "\n",
    "# plt.show()\n",
    "\n",
    "########################################\n",
    "# Break down by dataset\n",
    "########################################\n",
    "\n",
    "# Set up subplots.\n",
    "fig, axes = plt.subplots(nrows=1, ncols=2, sharey=True, sharex=False, figsize=(15,4))\n",
    "\n",
    "# Dataset 1: P18\n",
    "ax1 = plot_metric(\n",
    "    _df[_df.corpus==\"p18\"],\n",
    "    metric,\n",
    "    ax=axes[0],\n",
    "    ylabel=metric_label,\n",
    "    **kwargs\n",
    ")\n",
    "ax1.get_legend().remove()\n",
    "ax1.set_title(\"P18\")\n",
    "\n",
    "# Dataset 2: News\n",
    "ax2 = plot_metric(\n",
    "    _df[_df.corpus==\"news\"],\n",
    "    metric,\n",
    "    ax=axes[1],\n",
    "    ylabel=\"\",\n",
    "    **kwargs\n",
    ")\n",
    "ax2.set_title(\"News\")\n",
    "ax2.legend(loc='upper center', bbox_to_anchor=(-0.05, 1.3),\n",
    "          ncol=len(EVAL_TYPES))\n",
    "\n",
    "plt.subplots_adjust(wspace=0.1)\n",
    "# render(\"exp1_task_performance_bydataset.pdf\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e77cecf-78cf-4f30-bcb9-88e107c3466b",
   "metadata": {},
   "source": [
    "## Experiment 2: Word comparison"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "07f7f799",
   "metadata": {},
   "source": [
    "Updated"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "ba4d4266",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "d:\\Uni\\metalinguistic-prompting\\utils\\plotting.py:165: UserWarning: set_ticklabels() should only be used with a fixed number of ticks, i.e. after set_ticks() or using a FixedLocator.\n",
      "  ax.set_xticklabels(labels, rotation=25, ha=\"right\")\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = plot_metric(\n",
    "    RESULTS[\"exp2\"],\n",
    "    \"model_prefers_good_continuation\",\n",
    "    x=\"model_with_revision\",\n",
    "    hue=\"eval_type\",\n",
    "    xlabel=\"\",\n",
    "    ylabel=\"Accuracy\",\n",
    "    # legend_pos=\"outside\",\n",
    "    hue_order=EVAL_TYPES,\n",
    "    chance=0.5,\n",
    "    ylim=(0,1)\n",
    ")\n",
    "ax.get_legend().remove()\n",
    "plt.gcf().set_size_inches(*BARPLOT_SIZE)\n",
    "# render(\"exp2_task_performance.pdf\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a3f4408-4777-49d3-ac17-3561f5026b26",
   "metadata": {},
   "source": [
    "## Experiment 3b: Sentence comparison\n",
    "\n",
    "NOTE: It's simpler to generate the plots for Experiment 3b than the plots for Experiment 3a, so we'll do it first."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f6dabb44-5e78-49ea-b4ba-b366443116c9",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "d:\\Uni\\metalinguistic-prompting\\utils\\plotting.py:165: UserWarning: set_ticklabels() should only be used with a fixed number of ticks, i.e. after set_ticks() or using a FixedLocator.\n",
      "  ax.set_xticklabels(labels, rotation=25, ha=\"right\")\n",
      "d:\\Uni\\metalinguistic-prompting\\utils\\plotting.py:165: UserWarning: set_ticklabels() should only be used with a fixed number of ticks, i.e. after set_ticks() or using a FixedLocator.\n",
      "  ax.set_xticklabels(labels, rotation=25, ha=\"right\")\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "_df = RESULTS[\"exp3b\"]\n",
    "metric = \"model_prefers_good_sentence\"\n",
    "metric_label = \"Accuracy\"\n",
    "kwargs = dict(\n",
    "    x=\"model_with_revision\",\n",
    "    hue=\"eval_type\",\n",
    "    xlabel=\"\",\n",
    "    hue_order=EVAL_TYPES,\n",
    "    chance=0.5,\n",
    "    ylim=(0,1),\n",
    ")\n",
    "\n",
    "########################################\n",
    "# Average over datasets\n",
    "########################################\n",
    "\n",
    "# ax = plot_metric(_df, metric, ylabel=metric_label, **kwargs)\n",
    "# ax.get_legend().remove()\n",
    "# plt.gcf().set_size_inches(*BARPLOT_SIZE)\n",
    "# # render(\"exp3b_task_performance.pdf\")\n",
    "\n",
    "# plt.show()\n",
    "\n",
    "########################################\n",
    "# Break down by dataset\n",
    "########################################\n",
    "\n",
    "# Set up subplots.\n",
    "fig, axes = plt.subplots(nrows=1, ncols=2, sharey=True, sharex=False, figsize=(15,4))\n",
    "\n",
    "# Dataset 1: SyntaxGym\n",
    "ax1 = plot_metric(\n",
    "    _df[_df.corpus==\"syntaxgym\"],\n",
    "    metric,\n",
    "    ylabel=metric_label,\n",
    "    ax=axes[0],\n",
    "    **kwargs\n",
    ")\n",
    "ax1.get_legend().remove()\n",
    "ax1.set_title(\"SyntaxGym\")\n",
    "\n",
    "# Dataset 2: BLiMP\n",
    "ax2 = plot_metric(\n",
    "    _df[_df.corpus==\"blimp\"],\n",
    "    metric,\n",
    "    ylabel=\"\",\n",
    "    ax=axes[1],\n",
    "    **kwargs\n",
    ")\n",
    "ax2.set_title(\"BLiMP\")\n",
    "ax2.legend(loc='upper center', bbox_to_anchor=(-0.05, 1.3),\n",
    "          ncol=len(EVAL_TYPES))\n",
    "\n",
    "plt.subplots_adjust(wspace=0.1)\n",
    "# render(\"exp3b_task_performance_bydataset.pdf\")"
   ]
  }
 ],
 "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": "3.11.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}