{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "crazy-burden",
   "metadata": {
    "pycharm": {
     "name": "#%% Imports\n"
    }
   },
   "outputs": [],
   "source": [
    "from collections import defaultdict\n",
    "from pathlib import Path\n",
    "\n",
    "import numpy as np\n",
    "import torch\n",
    "import pytorch_lightning as pl\n",
    "import librosa\n",
    "import pandas as pd\n",
    "import variables as v\n",
    "import seaborn as sns\n",
    "from tqdm import tqdm\n",
    "from matplotlib import pyplot as plt\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ongoing-darkness",
   "metadata": {
    "pycharm": {
     "name": "#%% Variables\n"
    }
   },
   "outputs": [],
   "source": [
    "_ROOT = Path('..')\n",
    "output = Path('..') / Path('output')\n",
    "sr = 16000\n",
    "_root = v.CCS_Root\n",
    "roots = [v.CCS_Root, v.PRIMATES_Root]\n",
    "ext = 'pdf'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "friendly-tribute",
   "metadata": {
    "pycharm": {
     "name": "#%% Util Functions\n"
    }
   },
   "outputs": [],
   "source": [
    "def print_stats(data_option, mean_duration, std_duration, min_duration, max_duration):\n",
    "    print(f'For {data_option}; statistics are:')\n",
    "    print(f'Mean duration: {mean_duration:.3f}s\\tstd: {std_duration:.3f}s '\n",
    "          f'min: {min_duration:.3f}s\\t max: {max_duration:.3f}s')\n",
    "\n",
    "def show_and_save(data_option, title):\n",
    "    plt.tight_layout()\n",
    "    figure = plt.gcf()\n",
    "    plt.show()\n",
    "    figure.savefig(f'figures/{data_option}_{title.lower().replace(\" \", \"_\")}.{ext}', format=ext, )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "outputs": [
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001B[1;31m---------------------------------------------------------------------------\u001B[0m",
      "\u001B[1;31mKeyboardInterrupt\u001B[0m                         Traceback (most recent call last)",
      "\u001B[1;32m<ipython-input-4-fa304702b81b>\u001B[0m in \u001B[0;36m<module>\u001B[1;34m\u001B[0m\n\u001B[0;32m      7\u001B[0m         \u001B[0mdurations_per_class\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mdefaultdict\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mlist\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m      8\u001B[0m         \u001B[1;32mfor\u001B[0m \u001B[0midx\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mrow\u001B[0m \u001B[1;32min\u001B[0m \u001B[0mlab_csv\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0miterrows\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m----> 9\u001B[1;33m             \u001B[0mduration\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mlibrosa\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mget_duration\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfilename\u001B[0m\u001B[1;33m=\u001B[0m\u001B[0mstr\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mroot\u001B[0m \u001B[1;33m/\u001B[0m \u001B[1;34m'wav'\u001B[0m \u001B[1;33m/\u001B[0m \u001B[0mrow\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mfilename\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0msr\u001B[0m\u001B[1;33m=\u001B[0m\u001B[0msr\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m     10\u001B[0m             \u001B[0mdurations\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mappend\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mduration\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m     11\u001B[0m             \u001B[0mlabels\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mappend\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mrow\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mlabel\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32mc:\\users\\steff\\envs\\compare_21\\lib\\site-packages\\librosa\\core\\audio.py\u001B[0m in \u001B[0;36mget_duration\u001B[1;34m(y, sr, S, n_fft, hop_length, center, filename)\u001B[0m\n\u001B[0;32m    677\u001B[0m     \u001B[1;32mif\u001B[0m \u001B[0mfilename\u001B[0m \u001B[1;32mis\u001B[0m \u001B[1;32mnot\u001B[0m \u001B[1;32mNone\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    678\u001B[0m         \u001B[1;32mtry\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m--> 679\u001B[1;33m             \u001B[1;32mreturn\u001B[0m \u001B[0msf\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0minfo\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfilename\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mduration\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m    680\u001B[0m         \u001B[1;32mexcept\u001B[0m \u001B[0mRuntimeError\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    681\u001B[0m             \u001B[1;32mwith\u001B[0m \u001B[0maudioread\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0maudio_open\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfilename\u001B[0m\u001B[1;33m)\u001B[0m \u001B[1;32mas\u001B[0m \u001B[0mfdesc\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32mc:\\users\\steff\\envs\\compare_21\\lib\\site-packages\\soundfile.py\u001B[0m in \u001B[0;36minfo\u001B[1;34m(file, verbose)\u001B[0m\n\u001B[0;32m    436\u001B[0m         \u001B[0mWhether\u001B[0m \u001B[0mto\u001B[0m \u001B[0mprint\u001B[0m \u001B[0madditional\u001B[0m \u001B[0minformation\u001B[0m\u001B[1;33m.\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    437\u001B[0m     \"\"\"\n\u001B[1;32m--> 438\u001B[1;33m     \u001B[1;32mreturn\u001B[0m \u001B[0m_SoundFileInfo\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfile\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mverbose\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m    439\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    440\u001B[0m \u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32mc:\\users\\steff\\envs\\compare_21\\lib\\site-packages\\soundfile.py\u001B[0m in \u001B[0;36m__init__\u001B[1;34m(self, file, verbose)\u001B[0m\n\u001B[0;32m    381\u001B[0m     \u001B[1;32mdef\u001B[0m \u001B[0m__init__\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mself\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mfile\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mverbose\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    382\u001B[0m         \u001B[0mself\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mverbose\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mverbose\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m--> 383\u001B[1;33m         \u001B[1;32mwith\u001B[0m \u001B[0mSoundFile\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfile\u001B[0m\u001B[1;33m)\u001B[0m \u001B[1;32mas\u001B[0m \u001B[0mf\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m    384\u001B[0m             \u001B[0mself\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mname\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mf\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mname\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    385\u001B[0m             \u001B[0mself\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0msamplerate\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mf\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0msamplerate\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32mc:\\users\\steff\\envs\\compare_21\\lib\\site-packages\\soundfile.py\u001B[0m in \u001B[0;36m__init__\u001B[1;34m(self, file, mode, samplerate, channels, subtype, endian, format, closefd)\u001B[0m\n\u001B[0;32m    627\u001B[0m         self._info = _create_info_struct(file, mode, samplerate, channels,\n\u001B[0;32m    628\u001B[0m                                          format, subtype, endian)\n\u001B[1;32m--> 629\u001B[1;33m         \u001B[0mself\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0m_file\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mself\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0m_open\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfile\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mmode_int\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mclosefd\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m    630\u001B[0m         \u001B[1;32mif\u001B[0m \u001B[0mset\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mmode\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0missuperset\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;34m'r+'\u001B[0m\u001B[1;33m)\u001B[0m \u001B[1;32mand\u001B[0m \u001B[0mself\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mseekable\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m    631\u001B[0m             \u001B[1;31m# Move write position to 0 (like in Python file objects)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;32mc:\\users\\steff\\envs\\compare_21\\lib\\site-packages\\soundfile.py\u001B[0m in \u001B[0;36m_open\u001B[1;34m(self, file, mode_int, closefd)\u001B[0m\n\u001B[0;32m   1173\u001B[0m                 \u001B[1;32melse\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m   1174\u001B[0m                     \u001B[0mfile\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mfile\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mencode\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0m_sys\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0mgetfilesystemencoding\u001B[0m\u001B[1;33m(\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[1;32m-> 1175\u001B[1;33m             \u001B[0mfile_ptr\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0mopenfunction\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfile\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mmode_int\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mself\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0m_info\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0m\u001B[0;32m   1176\u001B[0m         \u001B[1;32melif\u001B[0m \u001B[0misinstance\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfile\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mint\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m:\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n\u001B[0;32m   1177\u001B[0m             \u001B[0mfile_ptr\u001B[0m \u001B[1;33m=\u001B[0m \u001B[0m_snd\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0msf_open_fd\u001B[0m\u001B[1;33m(\u001B[0m\u001B[0mfile\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mmode_int\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mself\u001B[0m\u001B[1;33m.\u001B[0m\u001B[0m_info\u001B[0m\u001B[1;33m,\u001B[0m \u001B[0mclosefd\u001B[0m\u001B[1;33m)\u001B[0m\u001B[1;33m\u001B[0m\u001B[1;33m\u001B[0m\u001B[0m\n",
      "\u001B[1;31mKeyboardInterrupt\u001B[0m: "
     ]
    }
   ],
   "source": [
    "data = defaultdict(dict)\n",
    "for root in roots:\n",
    "    for data_option in ['train', 'devel', 'test']:\n",
    "        lab_csv = pd.read_csv(root / 'lab' / f'{data_option}.csv')\n",
    "        labels = []\n",
    "        durations = []\n",
    "        durations_per_class = defaultdict(list)\n",
    "        for idx, row in lab_csv.iterrows():\n",
    "            duration = librosa.get_duration(filename=str(root / 'wav' / row.filename), sr=sr)\n",
    "            durations.append(duration)\n",
    "            labels.append(row.label)\n",
    "            durations_per_class[row.label].append(duration)\n",
    "        durations_per_class = dict(durations_per_class)\n",
    "        data[f'{root.name}_{data_option}'].update(durations=durations, labels=labels, durations_per_class=durations_per_class)\n",
    "\n",
    "data = dict(data)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% Prepare Data\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "for data_option in data.keys():\n",
    "    # For all files in data_option\n",
    "    durations = data[data_option]['durations']\n",
    "    print_stats(data_option, np.mean(durations), np.std(durations), np.min(durations), np.max(durations))\n",
    "\n",
    "    # Per class in data_option\n",
    "    durations_per_class = data[data_option]['durations_per_class']\n",
    "    for label in durations_per_class.keys():\n",
    "        cls_dur = durations_per_class[label]\n",
    "        print_stats(label, np.mean(cls_dur), np.std(cls_dur), np.min(cls_dur), np.max(cls_dur))\n",
    "    print('----------------------######################-----------------\\n')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n",
     "is_executing": true
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.style.use('default')\n",
    "sns.set_palette('Dark2')\n",
    "font_size = 16\n",
    "tex_fonts = {\n",
    "    # Use LaTeX to write all text\n",
    "    \"text.usetex\": True,\n",
    "    \"font.family\": \"serif\",\n",
    "    # Use 10pt font in plots, to match 10pt font in document\n",
    "    \"axes.labelsize\": font_size,\n",
    "    \"font.size\": font_size,\n",
    "    # Make the legend/label fonts a little smaller\n",
    "    \"legend.fontsize\": font_size-2,\n",
    "    \"xtick.labelsize\": font_size-2,\n",
    "    \"ytick.labelsize\": font_size-2\n",
    "}\n",
    "\n",
    "plt.rcParams.update(tex_fonts)\n",
    "\n",
    "Path('figures').mkdir(exist_ok=True)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% Seaborn Settings\n",
     "is_executing": true
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "for root in roots:\n",
    "    for data_option in ['train', 'devel']:\n",
    "        title = 'durations per class'\n",
    "        key = f'{root.name}_{data_option}'\n",
    "        test_key = f'{root.name}_test'\n",
    "\n",
    "        # Create the pandas DataFrame\n",
    "        df = pd.DataFrame().from_dict({key:val for key, val in data[key].items() if key!='durations_per_class'})\n",
    "        df_test = pd.DataFrame().from_dict({key:val for key, val in data[test_key].items() if key!='durations_per_class'})\n",
    "        df = df.append(df_test)\n",
    "\n",
    "        # Plotting\n",
    "        histplot = sns.histplot(\n",
    "            data=df, x=\"durations\", hue=\"labels\",\n",
    "            hue_order=sorted((*data[key]['durations_per_class'].keys(), '?')),\n",
    "            log_scale=False, element=\"step\", fill=False, bins='auto',\n",
    "            cumulative=False, stat=\"density\", common_norm=False,\n",
    "        )\n",
    "        #histplot.set(title=f'{data_option}-dataset {title}')\n",
    "\n",
    "        show_and_save(key, title)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% Histogram plotting for durations\n",
     "is_executing": true
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "for data_option in data.keys():\n",
    "    title = 'counts per class'\n",
    "    # Create the pandas DataFrame\n",
    "    df = pd.DataFrame().from_dict({key:val for key, val in data[data_option].items() if key!='durations_per_class'})\n",
    "    # Plotting\n",
    "    print(df.groupby('labels').count())\n",
    "    plot = sns.countplot(\n",
    "        data=df.sort_values('labels'), x='labels',# hue=\"labels\",\n",
    "        # hue_order=sorted(data[data_option]['durations_per_class'].keys()),\n",
    "\n",
    "    )\n",
    "    #plot.set(title=f'{data_option}-dataset {title}')\n",
    "\n",
    "    for p in plot.patches:\n",
    "        y_mod = 150 if 'CCS' in data_option else 300\n",
    "        plot.annotate(format(round(p.get_height()), '.0f'),\n",
    "                       (p.get_x() + p.get_width() / 2., p.get_height()),\n",
    "                       ha = 'center', va = 'center',\n",
    "                       size=12,\n",
    "                       xytext = (0, (-12) if p.get_height() > y_mod else (10)),\n",
    "                       textcoords = 'offset points')\n",
    "\n",
    "    show_and_save(data_option, title)\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% Histogram plotting for class Count\n",
     "is_executing": true
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "for data_option in data.keys():\n",
    "    title = f'KDE-plot'\n",
    "    # Create the pandas DataFrame\n",
    "    df = pd.DataFrame().from_dict({key:val for key, val in data[data_option].items() if key!='durations_per_class'})\n",
    "    # Plotting\n",
    "    kdeplot = sns.kdeplot(\n",
    "        data=df, x=\"durations\", hue=\"labels\",\n",
    "        hue_order=sorted(data[data_option]['durations_per_class'].keys()),\n",
    "        # log_scale=False, element=\"step\", fill=False,\n",
    "        # cumulative=False, stat=\"density\", common_norm=False,\n",
    "        # kde=True\n",
    "    )\n",
    "    # kdeplot.set(title=f'{data_option}-dataset {title}')\n",
    "\n",
    "    show_and_save(data_option, title)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% Histogram plotting for durations\n",
     "is_executing": true
    }
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.8.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}