notebooks/ismir/mushra.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import scipy as sp\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy import stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "mushra_file = \"mushra_Test-20230410104920096967.csv\"\n",
    "data = pd.read_csv(mushra_file, sep=\",\", header=1, parse_dates=True, skiprows=[2], usecols=lambda x: x not in [\"Unnamed: 0\", \"Unnamed: 2\", \"Unnamed: 34\"])\n",
    "data = data.rename(columns={\"Sound filename\": \"time\", \"Unnamed: 3\": \"Device\"})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of data: 33\n",
      "Number of data after filtering: 32\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0     2023-04-09 15:53:21\n",
       "1     2023-04-09 10:39:27\n",
       "2     2023-04-09 10:23:58\n",
       "3     2023-04-08 21:14:47\n",
       "4     2023-04-08 20:14:23\n",
       "5     2023-04-08 14:15:20\n",
       "6     2023-04-08 13:28:12\n",
       "7     2023-04-08 04:19:28\n",
       "8     2023-04-07 17:16:38\n",
       "9     2023-04-07 15:20:35\n",
       "10    2023-04-07 10:54:40\n",
       "11    2023-04-07 09:51:18\n",
       "12    2023-04-07 05:25:21\n",
       "13    2023-04-07 04:56:48\n",
       "14    2023-04-06 20:44:51\n",
       "15    2023-04-06 14:40:10\n",
       "16    2023-04-06 04:47:30\n",
       "17    2023-04-05 14:11:48\n",
       "18    2023-04-05 13:41:13\n",
       "19    2023-04-05 08:07:35\n",
       "20    2023-04-04 22:04:07\n",
       "21    2023-04-04 19:00:14\n",
       "22    2023-04-04 14:52:08\n",
       "23    2023-04-04 14:44:32\n",
       "24    2023-04-04 14:32:16\n",
       "25    2023-04-04 13:51:49\n",
       "26    2023-04-04 13:47:14\n",
       "27    2023-04-04 11:56:09\n",
       "29    2023-04-04 11:32:38\n",
       "30    2023-04-04 11:22:30\n",
       "31    2023-04-04 11:11:47\n",
       "32    2023-04-04 10:56:35\n",
       "Name: time, dtype: object"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(f\"Number of data: {len(data)}\")\n",
    "\n",
    "# filter participants whos responds contains \"Others\" in the device column\n",
    "filtered_data = data[data[\"Device\"].str.contains(\"Others\") == False]\n",
    "\n",
    "print(f\"Number of data after filtering: {len(filtered_data)}\")\n",
    "\n",
    "# filtered_data = filtered_data[:15]\n",
    "filtered_data.columns\n",
    "filtered_data[\"time\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "raw_data = np.empty((2, 5, 3, len(filtered_data)))\n",
    "idx0_name = [\"f1\", \"m1\"]\n",
    "idx1_name = [\"gt\", \"ddsp\", \"sawsing\", \"glottal_d\", \"pulse\"]\n",
    "idx2_name = list(range(1, 4))\n",
    "\n",
    "for i, j, k in np.ndindex(raw_data.shape[:3]):\n",
    "    # normalize data from [0, 100] to [1, 5] likert scale\n",
    "    raw_data[i, j, k] = filtered_data[f\"{idx1_name[j]}_{idx0_name[i]}_{idx2_name[k]}.wav\"].to_numpy() / 25 + 1\n",
    "raw_data = raw_data.reshape(2, 5, -1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[4.21208333 4.05125    2.61       3.43875    1.54958333]\n",
      " [4.08208333 3.57541667 2.3125     3.02708333 2.61416667]] [[0.14875946 0.15778246 0.16666004 0.15812764 0.09738725]\n",
      " [0.16807615 0.18284351 0.19359759 0.21413958 0.19548543]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 3000x1200 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# using the last axis as the likert scale, calculate the MOS mean and 95% confidence interval\n",
    "mos_mean = np.mean(raw_data, axis=-1)\n",
    "mos_std = np.std(raw_data, axis=-1)\n",
    "mos_ci = 1.96 * mos_std / np.sqrt(raw_data.shape[-1])\n",
    "\n",
    "print(mos_mean, mos_ci)\n",
    "sort_idx = np.argsort(mos_mean.mean(axis=0))\n",
    "plt_names = [idx1_name[i] for i in sort_idx]\n",
    "plt_mean = mos_mean[:, sort_idx]\n",
    "plt_ci = mos_ci[:, sort_idx]\n",
    "\n",
    "display_dict = {\n",
    "    \"gt\": \"Ref\",\n",
    "    \"ddsp\": \"DDSP\",\n",
    "    \"sawsing\": \"SawSing\",\n",
    "    \"glottal_d\": \"GOLF\",\n",
    "    \"pulse\": \"PULF\",\n",
    "}\n",
    "\n",
    "# plot the result using different color for each likert scale\n",
    "# sort the data by the mean of the first likert scale\n",
    "fig, ax = plt.subplots(1, 2, figsize=(10, 4), sharey=True, dpi=300)\n",
    "plt.rc(\"axes\", titlesize=24)\n",
    "plt.rc('axes', labelsize=20)\n",
    "plt.rc('ytick', labelsize=16)\n",
    "plt.rc('xtick', labelsize=18)\n",
    "for i in range(2):\n",
    "    ax[i].bar([display_dict[x] for x in plt_names], plt_mean[i], yerr=plt_ci[i],  align='center', ecolor='black', capsize=5)\n",
    "    ax[i].set_title(f\"{idx0_name[i]}\")\n",
    "    ax[i].set_ylim(ymin=1, ymax=5)\n",
    "    if i == 0:\n",
    "        ax[i].set_ylabel(\"Score\")\n",
    "    # ax[i].set_xlabel(\"Models\")\n",
    "    ax[i].grid(axis=\"y\")\n",
    "    ax[i].set_axisbelow(True)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ANOVA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0 0.14979945174550055\n",
      "[[0.00000000e+00 1.49799452e-01 5.18067178e-31 5.83851577e-11\n",
      "  3.72833914e-72]\n",
      " [0.00000000e+00 0.00000000e+00 8.65331355e-26 2.55990801e-07\n",
      "  2.98458516e-65]\n",
      " [0.00000000e+00 0.00000000e+00 0.00000000e+00 3.53890209e-11\n",
      "  3.02067618e-21]\n",
      " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
      "  3.64408583e-48]\n",
      " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
      "  0.00000000e+00]]\n"
     ]
    }
   ],
   "source": [
    "# do anova test for all the combinations and store them into a confusion matrix\n",
    "results = np.zeros((2, 5, 5))\n",
    "for i, j, k in np.ndindex(results.shape):\n",
    "    if j < k:\n",
    "        results[i, j, k] = stats.f_oneway(raw_data[i, j], raw_data[i, k])[1]\n",
    "\n",
    "print(results.min(), results.max())\n",
    "print(results[0])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Paired t-test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ttest_relResult(statistic=1.5786464580482067, pvalue=0.11774129420067447)\n",
      "Ttest_relResult(statistic=4.023028670737778, pvalue=0.00011532439635461517)\n"
     ]
    }
   ],
   "source": [
    "print(stats.ttest_rel(raw_data[0, 0], raw_data[0, 1]))\n",
    "print(stats.ttest_rel(raw_data[1, 0], raw_data[1, 1]))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Wilcoxon"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WilcoxonResult(statistic=4346.5, pvalue=0.9999999999999992)\n",
      "WilcoxonResult(statistic=1620.0, pvalue=0.021640322899261478)\n"
     ]
    }
   ],
   "source": [
    "print(stats.wilcoxon(raw_data[0, 2], raw_data[0, 4], alternative=\"less\"))\n",
    "print(stats.wilcoxon(raw_data[1, 2], raw_data[1, 4], alternative=\"less\"))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "torch2",
   "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": "3.9.12"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}