10. Batch Audio Features Extraction using Librosa.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "afbdd66b",
   "metadata": {},
   "source": [
    "# Acoustic Anomaly Detection based on Images with Azure AutoML for Images\n",
    "## 10. Batch Audio Features Extraction using Librosa"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ba65e974",
   "metadata": {},
   "source": [
    "<img src='https://github.com/retkowsky/CustomVision_Datasets/blob/master/anomaly.png?raw=true' width = 600>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c90d9701",
   "metadata": {},
   "source": [
    "Serge Retkowsky | serge.retkowsky@microsoft.com | 25-Oct-2022 | V3"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "86dbfbaa",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-info\"><b> Let's do the audio features extraction for all the files in a directory.</b></div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "9bc0e513",
   "metadata": {},
   "outputs": [],
   "source": [
    "#pip install xlwt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2c1fad78",
   "metadata": {},
   "outputs": [],
   "source": [
    "import datetime\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scipy\n",
    "import librosa\n",
    "import librosa.display\n",
    "import IPython\n",
    "import xlwt\n",
    "import sys\n",
    "\n",
    "from pathlib import Path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2584a4f7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'3.8.5 (default, Sep  4 2020, 07:30:14) \\n[GCC 7.3.0]'"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sys.version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6db6fc92",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2022-10-25 13:28:09.117711\n"
     ]
    }
   ],
   "source": [
    "now = datetime.datetime.now()\n",
    "print(now)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "a3bd17d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def count_files(mypath):\n",
    "    \"\"\"\n",
    "    Count files from a directory\n",
    "    \"\"\"\n",
    "    print(\"Total of files in\", mypath, \"=\",\n",
    "          sum(len(files) for _, _, files in os.walk(mypath)), '\\n')\n",
    "\n",
    "    for root, _, files in os.walk(mypath):\n",
    "        print(root, \":\", len(files), \"files\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce934a77",
   "metadata": {},
   "source": [
    "## 1. Directory to use to generate audio features for each file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4a1ac8da",
   "metadata": {},
   "outputs": [],
   "source": [
    "AUDIOFILES_DIR = \"audio_samples\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "837f8b78",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total of files in audio_samples = 122 \n",
      "\n",
      "audio_samples : 0 files\n",
      "audio_samples/KO : 61 files\n",
      "audio_samples/OK : 61 files\n"
     ]
    }
   ],
   "source": [
    "count_files(AUDIOFILES_DIR)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b21264b2",
   "metadata": {},
   "source": [
    "## 2. Directory to save the results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "4c291a33",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Directory not exists. Creating directory...  audio_features\n",
      "Done.\n"
     ]
    }
   ],
   "source": [
    "FEATURES_DIR = 'audio_features'\n",
    "\n",
    "if os.path.isdir(FEATURES_DIR) == True:\n",
    "    print(\"Directory\", FEATURES_DIR, \"exists. So let's use it.\")\n",
    "    \n",
    "if os.path.isdir(FEATURES_DIR) == False:\n",
    "    print(\"Directory not exists. Creating directory... \", FEATURES_DIR)\n",
    "    os.mkdir(FEATURES_DIR)\n",
    "    print(\"Done.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bdf87a63",
   "metadata": {},
   "source": [
    "## 3. Function to extract audio features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "dccb4eaf",
   "metadata": {},
   "outputs": [],
   "source": [
    "def audio_features_extraction(audio):\n",
    "    \"\"\"\n",
    "    Audio features extraction\n",
    "    \"\"\"\n",
    "    print(datetime.datetime.now(),\n",
    "          \"Generating audio features extraction from:\", audio)\n",
    "\n",
    "    # Getting some file infromations\n",
    "    basename = os.path.basename(audio)\n",
    "    basenamewithoutext = os.path.splitext(basename)[0]\n",
    "    extension = os.path.splitext(audiofilelocation)[1][1:]\n",
    "    fullfilename = audiofilelocation\n",
    "    dirsname = os.path.dirname(audio)\n",
    "    category = dirsname.split(\"/\", 1)[1]\n",
    "\n",
    "    basename_df = pd.DataFrame({'filename': [basename]})\n",
    "    basenamewithoutext_df = pd.DataFrame({'basename': [basenamewithoutext]})\n",
    "    ext_df = pd.DataFrame({'extension': [extension]})\n",
    "    dirsname_df = pd.DataFrame({'dirsname': [dirsname]})\n",
    "    fullfilename_df = pd.DataFrame({'fullfilename': [fullfilename]})\n",
    "    category_df = pd.DataFrame({'category': [category]})\n",
    "\n",
    "    # dataframe with filename and directory information\n",
    "    filename_df = pd.concat([\n",
    "        fullfilename_df, dirsname_df, category_df, basename_df,\n",
    "        basenamewithoutext_df, ext_df\n",
    "    ],\n",
    "                            axis=1)\n",
    "\n",
    "    # Loading sound file\n",
    "    y, sr = librosa.load(audio)\n",
    "\n",
    "    # Harmonic & percussive\n",
    "    y_harmonic, y_percussive = librosa.effects.hpss(y)\n",
    "    tempo, beat_frames = librosa.beat.beat_track(y=y_harmonic, sr=sr)\n",
    "\n",
    "    # Chroma\n",
    "    chroma = librosa.feature.chroma_cens(y=y_harmonic, sr=sr)\n",
    "    chroma_mean = np.mean(chroma, axis=1)\n",
    "    chroma_std = np.std(chroma, axis=1)\n",
    "    octave = ['C', 'C#', 'D', 'D#', 'E', 'F', 'F#', 'G', 'G#', 'A', 'A#', 'B']\n",
    "\n",
    "    chroma_df = pd.DataFrame()\n",
    "\n",
    "    for i in range(0, 12):\n",
    "        chroma_df['chroma_mean_' + str(i)] = chroma_mean[i]\n",
    "\n",
    "    for i in range(0, 12):\n",
    "        chroma_df['chroma_std_' + str(i)] = chroma_mean[i]\n",
    "\n",
    "    chroma_df.loc[0] = np.concatenate((chroma_mean, chroma_std), axis=0)\n",
    "\n",
    "    # MFCCS\n",
    "    mfccs = librosa.feature.mfcc(y=y_harmonic, sr=sr, n_mfcc=13)\n",
    "    mfccs_mean = np.mean(mfccs, axis=1)\n",
    "    mfccs_std = np.std(mfccs, axis=1)\n",
    "    coeffs = np.arange(0, 13)\n",
    "\n",
    "    mfccs_df = pd.DataFrame()\n",
    "\n",
    "    for i in range(0, 13):\n",
    "        mfccs_df['mfccs_mean_' + str(i)] = mfccs_mean[i]\n",
    "\n",
    "    for i in range(0, 13):\n",
    "        mfccs_df['mfccs_std_' + str(i)] = mfccs_mean[i]\n",
    "\n",
    "    mfccs_df.loc[0] = np.concatenate((mfccs_mean, mfccs_std), axis=0)\n",
    "\n",
    "    # Cent\n",
    "    cent = librosa.feature.spectral_centroid(y=y, sr=sr)\n",
    "    cent_mean = np.mean(cent)\n",
    "    cent_std = np.std(cent)\n",
    "    cent_skew = scipy.stats.skew(cent, axis=1)[0]\n",
    "\n",
    "    # Contrast\n",
    "    contrast = librosa.feature.spectral_contrast(y=y_harmonic, sr=sr)\n",
    "    contrast_mean = np.mean(contrast, axis=1)\n",
    "    contrast_std = np.std(contrast, axis=1)\n",
    "    conts = np.arange(0, 7)\n",
    "    contrast_df = pd.DataFrame()\n",
    "\n",
    "    # rolloff\n",
    "    rolloff = librosa.feature.spectral_rolloff(y=y, sr=sr)\n",
    "    rolloff_mean = np.mean(rolloff)\n",
    "    rolloff_std = np.std(rolloff)\n",
    "    rolloff_skew = scipy.stats.skew(rolloff, axis=1)[0]\n",
    "\n",
    "    spectral_df = pd.DataFrame()\n",
    "    collist = ['cent_mean', 'cent_std', 'cent_skew']\n",
    "\n",
    "    for i in range(0, 7):\n",
    "        collist.append('contrast_mean_' + str(i))\n",
    "\n",
    "    for i in range(0, 7):\n",
    "        collist.append('contrast_std_' + str(i))\n",
    "\n",
    "    collist = collist + ['rolloff_mean', 'rolloff_std', 'rolloff_skew']\n",
    "\n",
    "    for c in collist:\n",
    "        spectral_df[c] = 0\n",
    "\n",
    "    data = np.concatenate(\n",
    "        ([cent_mean, cent_std, cent_skew], contrast_mean, contrast_std,\n",
    "         [rolloff_mean, rolloff_std, rolloff_std]),\n",
    "        axis=0)\n",
    "    spectral_df.loc[0] = data\n",
    "\n",
    "    # zrate\n",
    "    zrate = librosa.feature.zero_crossing_rate(y_harmonic)\n",
    "    zrate_mean = np.mean(zrate)\n",
    "    zrate_std = np.std(zrate)\n",
    "    zrate_skew = scipy.stats.skew(zrate, axis=1)[0]\n",
    "    zrate_df = pd.DataFrame()\n",
    "    zrate_df['zrate_mean'] = 0\n",
    "    zrate_df['zrate_std'] = 0\n",
    "    zrate_df['zrate_skew'] = 0\n",
    "    zrate_df.loc[0] = [zrate_mean, zrate_std, zrate_skew]\n",
    "\n",
    "    # beat\n",
    "    beat_df = pd.DataFrame()\n",
    "    beat_df['tempo'] = tempo\n",
    "    beat_df.loc[0] = tempo\n",
    "\n",
    "    # Additional informations\n",
    "    sr_df = pd.DataFrame({'sampling_rate': [sr]})\n",
    "    totalsamples = np.size(y)\n",
    "    duration = totalsamples / sr\n",
    "    totalsamples_df = pd.DataFrame({'total_samples': [totalsamples]})\n",
    "    duration_df = pd.DataFrame({'duration': [duration]})\n",
    "\n",
    "    # Saving audiofeatures to a dataframe\n",
    "    audiofeatures_df = pd.concat(\n",
    "        (chroma_df, mfccs_df, spectral_df, zrate_df, beat_df), axis=1)\n",
    "\n",
    "    # Saving all results into a single dataframe\n",
    "    filesaudio_df = pd.concat(\n",
    "        [filename_df, sr_df, totalsamples_df, duration_df, audiofeatures_df],\n",
    "        axis=1)\n",
    "\n",
    "    # Export to CSV for each audio file\n",
    "    outputfile = FEATURES_DIR + '/' + basenamewithoutext + '_features.csv'\n",
    "    filesaudio_df.to_csv(outputfile)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e5231100",
   "metadata": {},
   "source": [
    "## 4. Let's run the function to extract audio features for each file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "788807c4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2022-10-25 13:29:13.723171 Audio Features extraction...\n",
      "\n",
      "2022-10-25 13:29:13.793069 Generating audio features extraction from: audio_samples/KO/machine_KO_1.wav\n",
      "2022-10-25 13:29:16.124523 Generating audio features extraction from: audio_samples/KO/machine_KO_10.wav\n",
      "2022-10-25 13:29:17.594290 Generating audio features extraction from: audio_samples/KO/machine_KO_11.wav\n",
      "2022-10-25 13:29:19.036441 Generating audio features extraction from: audio_samples/KO/machine_KO_12.wav\n",
      "2022-10-25 13:29:20.476144 Generating audio features extraction from: audio_samples/KO/machine_KO_13.wav\n",
      "2022-10-25 13:29:21.932754 Generating audio features extraction from: audio_samples/KO/machine_KO_14.wav\n",
      "2022-10-25 13:29:23.553014 Generating audio features extraction from: audio_samples/KO/machine_KO_15.wav\n",
      "2022-10-25 13:29:25.005511 Generating audio features extraction from: audio_samples/KO/machine_KO_16.wav\n",
      "2022-10-25 13:29:26.473732 Generating audio features extraction from: audio_samples/KO/machine_KO_17.wav\n",
      "2022-10-25 13:29:27.975041 Generating audio features extraction from: audio_samples/KO/machine_KO_18.wav\n",
      "2022-10-25 13:29:29.408561 Generating audio features extraction from: audio_samples/KO/machine_KO_19.wav\n",
      "2022-10-25 13:29:30.916549 Generating audio features extraction from: audio_samples/KO/machine_KO_2.wav\n",
      "2022-10-25 13:29:32.327475 Generating audio features extraction from: audio_samples/KO/machine_KO_20.wav\n",
      "2022-10-25 13:29:33.859465 Generating audio features extraction from: audio_samples/KO/machine_KO_21.wav\n",
      "2022-10-25 13:29:35.352720 Generating audio features extraction from: audio_samples/KO/machine_KO_22.wav\n",
      "2022-10-25 13:29:36.786266 Generating audio features extraction from: audio_samples/KO/machine_KO_23.wav\n",
      "2022-10-25 13:29:38.261880 Generating audio features extraction from: audio_samples/KO/machine_KO_24.wav\n",
      "2022-10-25 13:29:39.709728 Generating audio features extraction from: audio_samples/KO/machine_KO_25.wav\n",
      "2022-10-25 13:29:41.088735 Generating audio features extraction from: audio_samples/KO/machine_KO_26.wav\n",
      "2022-10-25 13:29:42.535882 Generating audio features extraction from: audio_samples/KO/machine_KO_27.wav\n",
      "2022-10-25 13:29:43.928595 Generating audio features extraction from: audio_samples/KO/machine_KO_28.wav\n",
      "2022-10-25 13:29:45.328377 Generating audio features extraction from: audio_samples/KO/machine_KO_29.wav\n",
      "2022-10-25 13:29:46.642196 Generating audio features extraction from: audio_samples/KO/machine_KO_3.wav\n",
      "2022-10-25 13:29:48.031924 Generating audio features extraction from: audio_samples/KO/machine_KO_30.wav\n",
      "2022-10-25 13:29:49.421363 Generating audio features extraction from: audio_samples/KO/machine_KO_31.wav\n",
      "2022-10-25 13:29:50.892313 Generating audio features extraction from: audio_samples/KO/machine_KO_32.wav\n",
      "2022-10-25 13:29:52.266570 Generating audio features extraction from: audio_samples/KO/machine_KO_33.wav\n",
      "2022-10-25 13:29:53.650306 Generating audio features extraction from: audio_samples/KO/machine_KO_34.wav\n",
      "2022-10-25 13:29:54.982156 Generating audio features extraction from: audio_samples/KO/machine_KO_35.wav\n",
      "2022-10-25 13:29:56.424865 Generating audio features extraction from: audio_samples/KO/machine_KO_36.wav\n",
      "2022-10-25 13:29:57.777033 Generating audio features extraction from: audio_samples/KO/machine_KO_37.wav\n",
      "2022-10-25 13:29:59.118262 Generating audio features extraction from: audio_samples/KO/machine_KO_38.wav\n",
      "2022-10-25 13:30:00.578664 Generating audio features extraction from: audio_samples/KO/machine_KO_39.wav\n",
      "2022-10-25 13:30:02.091193 Generating audio features extraction from: audio_samples/KO/machine_KO_4.wav\n",
      "2022-10-25 13:30:03.556134 Generating audio features extraction from: audio_samples/KO/machine_KO_40.wav\n",
      "2022-10-25 13:30:05.003945 Generating audio features extraction from: audio_samples/KO/machine_KO_41.wav\n",
      "2022-10-25 13:30:06.429342 Generating audio features extraction from: audio_samples/KO/machine_KO_42.wav\n",
      "2022-10-25 13:30:07.765388 Generating audio features extraction from: audio_samples/KO/machine_KO_43.wav\n",
      "2022-10-25 13:30:09.252365 Generating audio features extraction from: audio_samples/KO/machine_KO_44.wav\n",
      "2022-10-25 13:30:10.650004 Generating audio features extraction from: audio_samples/KO/machine_KO_45.wav\n",
      "2022-10-25 13:30:12.063670 Generating audio features extraction from: audio_samples/KO/machine_KO_46.wav\n",
      "2022-10-25 13:30:13.485601 Generating audio features extraction from: audio_samples/KO/machine_KO_47.wav\n",
      "2022-10-25 13:30:14.950026 Generating audio features extraction from: audio_samples/KO/machine_KO_48.wav\n",
      "2022-10-25 13:30:16.338932 Generating audio features extraction from: audio_samples/KO/machine_KO_49.wav\n",
      "2022-10-25 13:30:17.724718 Generating audio features extraction from: audio_samples/KO/machine_KO_5.wav\n",
      "2022-10-25 13:30:19.094705 Generating audio features extraction from: audio_samples/KO/machine_KO_50.wav\n",
      "2022-10-25 13:30:20.524129 Generating audio features extraction from: audio_samples/KO/machine_KO_51.wav\n",
      "2022-10-25 13:30:21.867599 Generating audio features extraction from: audio_samples/KO/machine_KO_52.wav\n",
      "2022-10-25 13:30:23.214242 Generating audio features extraction from: audio_samples/KO/machine_KO_53.wav\n",
      "2022-10-25 13:30:24.587948 Generating audio features extraction from: audio_samples/KO/machine_KO_54.wav\n",
      "2022-10-25 13:30:25.945240 Generating audio features extraction from: audio_samples/KO/machine_KO_55.wav\n",
      "2022-10-25 13:30:27.347132 Generating audio features extraction from: audio_samples/KO/machine_KO_56.wav\n",
      "2022-10-25 13:30:28.713823 Generating audio features extraction from: audio_samples/KO/machine_KO_57.wav\n",
      "2022-10-25 13:30:30.090518 Generating audio features extraction from: audio_samples/KO/machine_KO_58.wav\n",
      "2022-10-25 13:30:31.469975 Generating audio features extraction from: audio_samples/KO/machine_KO_59.wav\n",
      "2022-10-25 13:30:32.911039 Generating audio features extraction from: audio_samples/KO/machine_KO_6.wav\n",
      "2022-10-25 13:30:34.240046 Generating audio features extraction from: audio_samples/KO/machine_KO_60.wav\n",
      "2022-10-25 13:30:35.648970 Generating audio features extraction from: audio_samples/KO/machine_KO_61.wav\n",
      "2022-10-25 13:30:36.548888 Generating audio features extraction from: audio_samples/KO/machine_KO_7.wav\n",
      "2022-10-25 13:30:37.847345 Generating audio features extraction from: audio_samples/KO/machine_KO_8.wav\n",
      "2022-10-25 13:30:39.203189 Generating audio features extraction from: audio_samples/KO/machine_KO_9.wav\n",
      "2022-10-25 13:30:40.580234 Generating audio features extraction from: audio_samples/OK/machine_OK_1.wav\n",
      "2022-10-25 13:30:41.961410 Generating audio features extraction from: audio_samples/OK/machine_OK_10.wav\n",
      "2022-10-25 13:30:43.358979 Generating audio features extraction from: audio_samples/OK/machine_OK_11.wav\n",
      "2022-10-25 13:30:44.828176 Generating audio features extraction from: audio_samples/OK/machine_OK_12.wav\n",
      "2022-10-25 13:30:46.386292 Generating audio features extraction from: audio_samples/OK/machine_OK_13.wav\n",
      "2022-10-25 13:30:47.780832 Generating audio features extraction from: audio_samples/OK/machine_OK_14.wav\n",
      "2022-10-25 13:30:49.194559 Generating audio features extraction from: audio_samples/OK/machine_OK_15.wav\n",
      "2022-10-25 13:30:50.604817 Generating audio features extraction from: audio_samples/OK/machine_OK_16.wav\n",
      "2022-10-25 13:30:52.006663 Generating audio features extraction from: audio_samples/OK/machine_OK_17.wav\n",
      "2022-10-25 13:30:53.353410 Generating audio features extraction from: audio_samples/OK/machine_OK_18.wav\n",
      "2022-10-25 13:30:54.773380 Generating audio features extraction from: audio_samples/OK/machine_OK_19.wav\n",
      "2022-10-25 13:30:56.151341 Generating audio features extraction from: audio_samples/OK/machine_OK_2.wav\n",
      "2022-10-25 13:30:57.578254 Generating audio features extraction from: audio_samples/OK/machine_OK_20.wav\n",
      "2022-10-25 13:30:58.903550 Generating audio features extraction from: audio_samples/OK/machine_OK_21.wav\n",
      "2022-10-25 13:31:00.249558 Generating audio features extraction from: audio_samples/OK/machine_OK_22.wav\n",
      "2022-10-25 13:31:01.680619 Generating audio features extraction from: audio_samples/OK/machine_OK_23.wav\n",
      "2022-10-25 13:31:03.075081 Generating audio features extraction from: audio_samples/OK/machine_OK_24.wav\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2022-10-25 13:31:04.453705 Generating audio features extraction from: audio_samples/OK/machine_OK_25.wav\n",
      "2022-10-25 13:31:05.943796 Generating audio features extraction from: audio_samples/OK/machine_OK_26.wav\n",
      "2022-10-25 13:31:07.342643 Generating audio features extraction from: audio_samples/OK/machine_OK_27.wav\n",
      "2022-10-25 13:31:08.763243 Generating audio features extraction from: audio_samples/OK/machine_OK_28.wav\n",
      "2022-10-25 13:31:10.128919 Generating audio features extraction from: audio_samples/OK/machine_OK_29.wav\n",
      "2022-10-25 13:31:11.460132 Generating audio features extraction from: audio_samples/OK/machine_OK_3.wav\n",
      "2022-10-25 13:31:12.812715 Generating audio features extraction from: audio_samples/OK/machine_OK_30.wav\n",
      "2022-10-25 13:31:14.280133 Generating audio features extraction from: audio_samples/OK/machine_OK_31.wav\n",
      "2022-10-25 13:31:15.617847 Generating audio features extraction from: audio_samples/OK/machine_OK_32.wav\n",
      "2022-10-25 13:31:17.020286 Generating audio features extraction from: audio_samples/OK/machine_OK_33.wav\n",
      "2022-10-25 13:31:18.419720 Generating audio features extraction from: audio_samples/OK/machine_OK_34.wav\n",
      "2022-10-25 13:31:19.738787 Generating audio features extraction from: audio_samples/OK/machine_OK_35.wav\n",
      "2022-10-25 13:31:21.115602 Generating audio features extraction from: audio_samples/OK/machine_OK_36.wav\n",
      "2022-10-25 13:31:22.470242 Generating audio features extraction from: audio_samples/OK/machine_OK_37.wav\n",
      "2022-10-25 13:31:23.818787 Generating audio features extraction from: audio_samples/OK/machine_OK_38.wav\n",
      "2022-10-25 13:31:25.181494 Generating audio features extraction from: audio_samples/OK/machine_OK_39.wav\n",
      "2022-10-25 13:31:26.537239 Generating audio features extraction from: audio_samples/OK/machine_OK_4.wav\n",
      "2022-10-25 13:31:27.909502 Generating audio features extraction from: audio_samples/OK/machine_OK_40.wav\n",
      "2022-10-25 13:31:29.272594 Generating audio features extraction from: audio_samples/OK/machine_OK_41.wav\n",
      "2022-10-25 13:31:30.604251 Generating audio features extraction from: audio_samples/OK/machine_OK_42.wav\n",
      "2022-10-25 13:31:31.906213 Generating audio features extraction from: audio_samples/OK/machine_OK_43.wav\n",
      "2022-10-25 13:31:33.244950 Generating audio features extraction from: audio_samples/OK/machine_OK_44.wav\n",
      "2022-10-25 13:31:34.724655 Generating audio features extraction from: audio_samples/OK/machine_OK_45.wav\n",
      "2022-10-25 13:31:36.023364 Generating audio features extraction from: audio_samples/OK/machine_OK_46.wav\n",
      "2022-10-25 13:31:37.343845 Generating audio features extraction from: audio_samples/OK/machine_OK_47.wav\n",
      "2022-10-25 13:31:38.649017 Generating audio features extraction from: audio_samples/OK/machine_OK_48.wav\n",
      "2022-10-25 13:31:39.972108 Generating audio features extraction from: audio_samples/OK/machine_OK_49.wav\n",
      "2022-10-25 13:31:41.289747 Generating audio features extraction from: audio_samples/OK/machine_OK_5.wav\n",
      "2022-10-25 13:31:42.594693 Generating audio features extraction from: audio_samples/OK/machine_OK_50.wav\n",
      "2022-10-25 13:31:43.905571 Generating audio features extraction from: audio_samples/OK/machine_OK_51.wav\n",
      "2022-10-25 13:31:45.264003 Generating audio features extraction from: audio_samples/OK/machine_OK_52.wav\n",
      "2022-10-25 13:31:46.618590 Generating audio features extraction from: audio_samples/OK/machine_OK_53.wav\n",
      "2022-10-25 13:31:47.965901 Generating audio features extraction from: audio_samples/OK/machine_OK_54.wav\n",
      "2022-10-25 13:31:49.280571 Generating audio features extraction from: audio_samples/OK/machine_OK_55.wav\n",
      "2022-10-25 13:31:50.627593 Generating audio features extraction from: audio_samples/OK/machine_OK_56.wav\n",
      "2022-10-25 13:31:51.982516 Generating audio features extraction from: audio_samples/OK/machine_OK_57.wav\n",
      "2022-10-25 13:31:53.355099 Generating audio features extraction from: audio_samples/OK/machine_OK_58.wav\n",
      "2022-10-25 13:31:54.735534 Generating audio features extraction from: audio_samples/OK/machine_OK_59.wav\n",
      "2022-10-25 13:31:56.011507 Generating audio features extraction from: audio_samples/OK/machine_OK_6.wav\n",
      "2022-10-25 13:31:57.385930 Generating audio features extraction from: audio_samples/OK/machine_OK_60.wav\n",
      "2022-10-25 13:31:58.733615 Generating audio features extraction from: audio_samples/OK/machine_OK_61.wav\n",
      "2022-10-25 13:31:59.607862 Generating audio features extraction from: audio_samples/OK/machine_OK_7.wav\n",
      "2022-10-25 13:32:00.944217 Generating audio features extraction from: audio_samples/OK/machine_OK_8.wav\n",
      "2022-10-25 13:32:02.439710 Generating audio features extraction from: audio_samples/OK/machine_OK_9.wav\n",
      "\n",
      " 2022-10-25 13:32:03.779580 End of job. Processing time: 0:02:50.056409\n"
     ]
    }
   ],
   "source": [
    "start = datetime.datetime.now()\n",
    "\n",
    "print(start, 'Audio Features extraction...\\n')\n",
    "\n",
    "for root, dirs, files in os.walk(AUDIOFILES_DIR):\n",
    "    for audiofile in files:\n",
    "        audiofilelocation = root + '/' + audiofile\n",
    "        audio_features_extraction(audiofilelocation)\n",
    "\n",
    "end = datetime.datetime.now()\n",
    "print('\\n', end, 'End of job. Processing time:', end - start)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "df956cd4",
   "metadata": {},
   "source": [
    "## 5. Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ed0cb444",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total of files in audio_features = 122 \n",
      "\n",
      "audio_features : 122 files\n"
     ]
    }
   ],
   "source": [
    "count_files(FEATURES_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "cbfe5da8",
   "metadata": {},
   "outputs": [],
   "source": [
    "os.chdir(FEATURES_DIR)\n",
    "\n",
    "filepaths = [f for f in os.listdir() if f.endswith('.csv')] \n",
    "\n",
    "# Taking all the CSV files into a single dataframe\n",
    "audiofeaturesdf = pd.concat(map(pd.read_csv, filepaths))\n",
    "del audiofeaturesdf['Unnamed: 0']  # Cleaning the df\n",
    "audiofeaturesdf = audiofeaturesdf.reset_index(drop=True)  # Clean index\n",
    "\n",
    "os.chdir('..')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "916537d0",
   "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>fullfilename</th>\n",
       "      <th>dirsname</th>\n",
       "      <th>category</th>\n",
       "      <th>filename</th>\n",
       "      <th>basename</th>\n",
       "      <th>extension</th>\n",
       "      <th>sampling_rate</th>\n",
       "      <th>total_samples</th>\n",
       "      <th>duration</th>\n",
       "      <th>chroma_mean_0</th>\n",
       "      <th>...</th>\n",
       "      <th>contrast_std_4</th>\n",
       "      <th>contrast_std_5</th>\n",
       "      <th>contrast_std_6</th>\n",
       "      <th>rolloff_mean</th>\n",
       "      <th>rolloff_std</th>\n",
       "      <th>rolloff_skew</th>\n",
       "      <th>zrate_mean</th>\n",
       "      <th>zrate_std</th>\n",
       "      <th>zrate_skew</th>\n",
       "      <th>tempo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>audio_samples/KO/machine_KO_10.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_10.wav</td>\n",
       "      <td>machine_KO_10</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.207503</td>\n",
       "      <td>...</td>\n",
       "      <td>1.687127</td>\n",
       "      <td>1.348918</td>\n",
       "      <td>4.272936</td>\n",
       "      <td>4407.604253</td>\n",
       "      <td>414.652790</td>\n",
       "      <td>414.652790</td>\n",
       "      <td>0.041735</td>\n",
       "      <td>0.007198</td>\n",
       "      <td>0.174384</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>audio_samples/KO/machine_KO_11.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_11.wav</td>\n",
       "      <td>machine_KO_11</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.254620</td>\n",
       "      <td>...</td>\n",
       "      <td>1.659826</td>\n",
       "      <td>1.223536</td>\n",
       "      <td>4.400989</td>\n",
       "      <td>4501.437681</td>\n",
       "      <td>387.931940</td>\n",
       "      <td>387.931940</td>\n",
       "      <td>0.043779</td>\n",
       "      <td>0.006834</td>\n",
       "      <td>-0.067536</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>audio_samples/KO/machine_KO_12.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_12.wav</td>\n",
       "      <td>machine_KO_12</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.209314</td>\n",
       "      <td>...</td>\n",
       "      <td>1.754141</td>\n",
       "      <td>1.625297</td>\n",
       "      <td>3.865849</td>\n",
       "      <td>4504.147157</td>\n",
       "      <td>507.148701</td>\n",
       "      <td>507.148701</td>\n",
       "      <td>0.044434</td>\n",
       "      <td>0.008300</td>\n",
       "      <td>0.590730</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>audio_samples/KO/machine_KO_13.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_13.wav</td>\n",
       "      <td>machine_KO_13</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.256801</td>\n",
       "      <td>...</td>\n",
       "      <td>1.821594</td>\n",
       "      <td>1.314911</td>\n",
       "      <td>4.268477</td>\n",
       "      <td>4438.014555</td>\n",
       "      <td>488.818772</td>\n",
       "      <td>488.818772</td>\n",
       "      <td>0.044020</td>\n",
       "      <td>0.007560</td>\n",
       "      <td>0.081320</td>\n",
       "      <td>129.199219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>audio_samples/KO/machine_KO_14.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_14.wav</td>\n",
       "      <td>machine_KO_14</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.260920</td>\n",
       "      <td>...</td>\n",
       "      <td>1.809934</td>\n",
       "      <td>1.436137</td>\n",
       "      <td>4.342983</td>\n",
       "      <td>4489.744154</td>\n",
       "      <td>418.799064</td>\n",
       "      <td>418.799064</td>\n",
       "      <td>0.044665</td>\n",
       "      <td>0.007445</td>\n",
       "      <td>0.094036</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>audio_samples/KO/machine_KO_15.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_15.wav</td>\n",
       "      <td>machine_KO_15</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.250668</td>\n",
       "      <td>...</td>\n",
       "      <td>1.590028</td>\n",
       "      <td>1.412233</td>\n",
       "      <td>4.306663</td>\n",
       "      <td>4621.260687</td>\n",
       "      <td>411.167343</td>\n",
       "      <td>411.167343</td>\n",
       "      <td>0.046744</td>\n",
       "      <td>0.007470</td>\n",
       "      <td>0.163075</td>\n",
       "      <td>172.265625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>audio_samples/KO/machine_KO_16.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_16.wav</td>\n",
       "      <td>machine_KO_16</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.212568</td>\n",
       "      <td>...</td>\n",
       "      <td>1.730367</td>\n",
       "      <td>1.322294</td>\n",
       "      <td>4.497835</td>\n",
       "      <td>4554.664622</td>\n",
       "      <td>477.516466</td>\n",
       "      <td>477.516466</td>\n",
       "      <td>0.044503</td>\n",
       "      <td>0.008391</td>\n",
       "      <td>-0.116057</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>audio_samples/KO/machine_KO_17.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_17.wav</td>\n",
       "      <td>machine_KO_17</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.254056</td>\n",
       "      <td>...</td>\n",
       "      <td>1.723347</td>\n",
       "      <td>1.283012</td>\n",
       "      <td>4.359413</td>\n",
       "      <td>4501.116822</td>\n",
       "      <td>447.835210</td>\n",
       "      <td>447.835210</td>\n",
       "      <td>0.040455</td>\n",
       "      <td>0.007335</td>\n",
       "      <td>0.323728</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>audio_samples/KO/machine_KO_18.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_18.wav</td>\n",
       "      <td>machine_KO_18</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.295805</td>\n",
       "      <td>...</td>\n",
       "      <td>1.734805</td>\n",
       "      <td>1.301398</td>\n",
       "      <td>4.176879</td>\n",
       "      <td>4621.367640</td>\n",
       "      <td>466.367128</td>\n",
       "      <td>466.367128</td>\n",
       "      <td>0.045609</td>\n",
       "      <td>0.008622</td>\n",
       "      <td>-0.160810</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>audio_samples/KO/machine_KO_19.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_19.wav</td>\n",
       "      <td>machine_KO_19</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.204857</td>\n",
       "      <td>...</td>\n",
       "      <td>1.736602</td>\n",
       "      <td>1.416512</td>\n",
       "      <td>4.189162</td>\n",
       "      <td>4753.205033</td>\n",
       "      <td>359.446599</td>\n",
       "      <td>359.446599</td>\n",
       "      <td>0.046316</td>\n",
       "      <td>0.008715</td>\n",
       "      <td>0.307786</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 83 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                         fullfilename          dirsname category  \\\n",
       "0  audio_samples/KO/machine_KO_10.wav  audio_samples/KO       KO   \n",
       "1  audio_samples/KO/machine_KO_11.wav  audio_samples/KO       KO   \n",
       "2  audio_samples/KO/machine_KO_12.wav  audio_samples/KO       KO   \n",
       "3  audio_samples/KO/machine_KO_13.wav  audio_samples/KO       KO   \n",
       "4  audio_samples/KO/machine_KO_14.wav  audio_samples/KO       KO   \n",
       "5  audio_samples/KO/machine_KO_15.wav  audio_samples/KO       KO   \n",
       "6  audio_samples/KO/machine_KO_16.wav  audio_samples/KO       KO   \n",
       "7  audio_samples/KO/machine_KO_17.wav  audio_samples/KO       KO   \n",
       "8  audio_samples/KO/machine_KO_18.wav  audio_samples/KO       KO   \n",
       "9  audio_samples/KO/machine_KO_19.wav  audio_samples/KO       KO   \n",
       "\n",
       "            filename       basename extension  sampling_rate  total_samples  \\\n",
       "0  machine_KO_10.wav  machine_KO_10       wav          22050         154350   \n",
       "1  machine_KO_11.wav  machine_KO_11       wav          22050         154350   \n",
       "2  machine_KO_12.wav  machine_KO_12       wav          22050         154350   \n",
       "3  machine_KO_13.wav  machine_KO_13       wav          22050         154350   \n",
       "4  machine_KO_14.wav  machine_KO_14       wav          22050         154350   \n",
       "5  machine_KO_15.wav  machine_KO_15       wav          22050         154350   \n",
       "6  machine_KO_16.wav  machine_KO_16       wav          22050         154350   \n",
       "7  machine_KO_17.wav  machine_KO_17       wav          22050         154350   \n",
       "8  machine_KO_18.wav  machine_KO_18       wav          22050         154350   \n",
       "9  machine_KO_19.wav  machine_KO_19       wav          22050         154350   \n",
       "\n",
       "   duration  chroma_mean_0  ...  contrast_std_4  contrast_std_5  \\\n",
       "0       7.0       0.207503  ...        1.687127        1.348918   \n",
       "1       7.0       0.254620  ...        1.659826        1.223536   \n",
       "2       7.0       0.209314  ...        1.754141        1.625297   \n",
       "3       7.0       0.256801  ...        1.821594        1.314911   \n",
       "4       7.0       0.260920  ...        1.809934        1.436137   \n",
       "5       7.0       0.250668  ...        1.590028        1.412233   \n",
       "6       7.0       0.212568  ...        1.730367        1.322294   \n",
       "7       7.0       0.254056  ...        1.723347        1.283012   \n",
       "8       7.0       0.295805  ...        1.734805        1.301398   \n",
       "9       7.0       0.204857  ...        1.736602        1.416512   \n",
       "\n",
       "   contrast_std_6  rolloff_mean  rolloff_std  rolloff_skew  zrate_mean  \\\n",
       "0        4.272936   4407.604253   414.652790    414.652790    0.041735   \n",
       "1        4.400989   4501.437681   387.931940    387.931940    0.043779   \n",
       "2        3.865849   4504.147157   507.148701    507.148701    0.044434   \n",
       "3        4.268477   4438.014555   488.818772    488.818772    0.044020   \n",
       "4        4.342983   4489.744154   418.799064    418.799064    0.044665   \n",
       "5        4.306663   4621.260687   411.167343    411.167343    0.046744   \n",
       "6        4.497835   4554.664622   477.516466    477.516466    0.044503   \n",
       "7        4.359413   4501.116822   447.835210    447.835210    0.040455   \n",
       "8        4.176879   4621.367640   466.367128    466.367128    0.045609   \n",
       "9        4.189162   4753.205033   359.446599    359.446599    0.046316   \n",
       "\n",
       "   zrate_std  zrate_skew       tempo  \n",
       "0   0.007198    0.174384   89.102909  \n",
       "1   0.006834   -0.067536   89.102909  \n",
       "2   0.008300    0.590730   89.102909  \n",
       "3   0.007560    0.081320  129.199219  \n",
       "4   0.007445    0.094036   89.102909  \n",
       "5   0.007470    0.163075  172.265625  \n",
       "6   0.008391   -0.116057   89.102909  \n",
       "7   0.007335    0.323728   89.102909  \n",
       "8   0.008622   -0.160810   89.102909  \n",
       "9   0.008715    0.307786   89.102909  \n",
       "\n",
       "[10 rows x 83 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "audiofeaturesdf.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "57fa2581",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(122, 83)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "audiofeaturesdf.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "61c510f6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['fullfilename', 'dirsname', 'category', 'filename', 'basename',\n",
       "       'extension', 'sampling_rate', 'total_samples', 'duration',\n",
       "       'chroma_mean_0', 'chroma_mean_1', 'chroma_mean_2', 'chroma_mean_3',\n",
       "       'chroma_mean_4', 'chroma_mean_5', 'chroma_mean_6', 'chroma_mean_7',\n",
       "       'chroma_mean_8', 'chroma_mean_9', 'chroma_mean_10', 'chroma_mean_11',\n",
       "       'chroma_std_0', 'chroma_std_1', 'chroma_std_2', 'chroma_std_3',\n",
       "       'chroma_std_4', 'chroma_std_5', 'chroma_std_6', 'chroma_std_7',\n",
       "       'chroma_std_8', 'chroma_std_9', 'chroma_std_10', 'chroma_std_11',\n",
       "       'mfccs_mean_0', 'mfccs_mean_1', 'mfccs_mean_2', 'mfccs_mean_3',\n",
       "       'mfccs_mean_4', 'mfccs_mean_5', 'mfccs_mean_6', 'mfccs_mean_7',\n",
       "       'mfccs_mean_8', 'mfccs_mean_9', 'mfccs_mean_10', 'mfccs_mean_11',\n",
       "       'mfccs_mean_12', 'mfccs_std_0', 'mfccs_std_1', 'mfccs_std_2',\n",
       "       'mfccs_std_3', 'mfccs_std_4', 'mfccs_std_5', 'mfccs_std_6',\n",
       "       'mfccs_std_7', 'mfccs_std_8', 'mfccs_std_9', 'mfccs_std_10',\n",
       "       'mfccs_std_11', 'mfccs_std_12', 'cent_mean', 'cent_std', 'cent_skew',\n",
       "       'contrast_mean_0', 'contrast_mean_1', 'contrast_mean_2',\n",
       "       'contrast_mean_3', 'contrast_mean_4', 'contrast_mean_5',\n",
       "       'contrast_mean_6', 'contrast_std_0', 'contrast_std_1', 'contrast_std_2',\n",
       "       'contrast_std_3', 'contrast_std_4', 'contrast_std_5', 'contrast_std_6',\n",
       "       'rolloff_mean', 'rolloff_std', 'rolloff_skew', 'zrate_mean',\n",
       "       'zrate_std', 'zrate_skew', 'tempo'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "audiofeaturesdf.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "372cfcff",
   "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>sampling_rate</th>\n",
       "      <th>total_samples</th>\n",
       "      <th>duration</th>\n",
       "      <th>chroma_mean_0</th>\n",
       "      <th>chroma_mean_1</th>\n",
       "      <th>chroma_mean_2</th>\n",
       "      <th>chroma_mean_3</th>\n",
       "      <th>chroma_mean_4</th>\n",
       "      <th>chroma_mean_5</th>\n",
       "      <th>chroma_mean_6</th>\n",
       "      <th>...</th>\n",
       "      <th>contrast_std_4</th>\n",
       "      <th>contrast_std_5</th>\n",
       "      <th>contrast_std_6</th>\n",
       "      <th>rolloff_mean</th>\n",
       "      <th>rolloff_std</th>\n",
       "      <th>rolloff_skew</th>\n",
       "      <th>zrate_mean</th>\n",
       "      <th>zrate_std</th>\n",
       "      <th>zrate_skew</th>\n",
       "      <th>tempo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>122.0</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>22050.0</td>\n",
       "      <td>153183.885246</td>\n",
       "      <td>6.947115</td>\n",
       "      <td>0.256014</td>\n",
       "      <td>0.096849</td>\n",
       "      <td>0.018883</td>\n",
       "      <td>0.005733</td>\n",
       "      <td>0.035318</td>\n",
       "      <td>0.214208</td>\n",
       "      <td>0.240684</td>\n",
       "      <td>...</td>\n",
       "      <td>1.707834</td>\n",
       "      <td>1.302730</td>\n",
       "      <td>4.088079</td>\n",
       "      <td>4664.787692</td>\n",
       "      <td>420.106882</td>\n",
       "      <td>420.106882</td>\n",
       "      <td>0.048327</td>\n",
       "      <td>0.008571</td>\n",
       "      <td>0.076895</td>\n",
       "      <td>108.375900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.0</td>\n",
       "      <td>9069.934393</td>\n",
       "      <td>0.411335</td>\n",
       "      <td>0.031143</td>\n",
       "      <td>0.038376</td>\n",
       "      <td>0.012205</td>\n",
       "      <td>0.005535</td>\n",
       "      <td>0.029643</td>\n",
       "      <td>0.031075</td>\n",
       "      <td>0.047745</td>\n",
       "      <td>...</td>\n",
       "      <td>0.088822</td>\n",
       "      <td>0.075221</td>\n",
       "      <td>0.247915</td>\n",
       "      <td>129.391998</td>\n",
       "      <td>68.570409</td>\n",
       "      <td>68.570409</td>\n",
       "      <td>0.003986</td>\n",
       "      <td>0.001031</td>\n",
       "      <td>0.216112</td>\n",
       "      <td>33.200135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>22050.0</td>\n",
       "      <td>83217.000000</td>\n",
       "      <td>3.774014</td>\n",
       "      <td>0.188875</td>\n",
       "      <td>0.004705</td>\n",
       "      <td>0.003091</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.134658</td>\n",
       "      <td>0.130089</td>\n",
       "      <td>...</td>\n",
       "      <td>1.508113</td>\n",
       "      <td>1.155480</td>\n",
       "      <td>3.485877</td>\n",
       "      <td>4346.819300</td>\n",
       "      <td>269.640466</td>\n",
       "      <td>269.640466</td>\n",
       "      <td>0.039852</td>\n",
       "      <td>0.006078</td>\n",
       "      <td>-0.363726</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>22050.0</td>\n",
       "      <td>154350.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.225710</td>\n",
       "      <td>0.067572</td>\n",
       "      <td>0.009924</td>\n",
       "      <td>0.001502</td>\n",
       "      <td>0.013512</td>\n",
       "      <td>0.193511</td>\n",
       "      <td>0.212933</td>\n",
       "      <td>...</td>\n",
       "      <td>1.656098</td>\n",
       "      <td>1.249198</td>\n",
       "      <td>3.939090</td>\n",
       "      <td>4568.060485</td>\n",
       "      <td>378.114109</td>\n",
       "      <td>378.114109</td>\n",
       "      <td>0.045546</td>\n",
       "      <td>0.007744</td>\n",
       "      <td>-0.084345</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>22050.0</td>\n",
       "      <td>154350.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.257555</td>\n",
       "      <td>0.102908</td>\n",
       "      <td>0.014943</td>\n",
       "      <td>0.003994</td>\n",
       "      <td>0.028536</td>\n",
       "      <td>0.210491</td>\n",
       "      <td>0.241342</td>\n",
       "      <td>...</td>\n",
       "      <td>1.705757</td>\n",
       "      <td>1.300551</td>\n",
       "      <td>4.137765</td>\n",
       "      <td>4659.638987</td>\n",
       "      <td>414.565745</td>\n",
       "      <td>414.565745</td>\n",
       "      <td>0.048476</td>\n",
       "      <td>0.008630</td>\n",
       "      <td>0.065217</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>22050.0</td>\n",
       "      <td>154350.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.283674</td>\n",
       "      <td>0.120758</td>\n",
       "      <td>0.023548</td>\n",
       "      <td>0.008489</td>\n",
       "      <td>0.048341</td>\n",
       "      <td>0.233583</td>\n",
       "      <td>0.259929</td>\n",
       "      <td>...</td>\n",
       "      <td>1.769236</td>\n",
       "      <td>1.344990</td>\n",
       "      <td>4.264405</td>\n",
       "      <td>4774.292598</td>\n",
       "      <td>465.569150</td>\n",
       "      <td>465.569150</td>\n",
       "      <td>0.050921</td>\n",
       "      <td>0.009270</td>\n",
       "      <td>0.223775</td>\n",
       "      <td>123.815918</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>22050.0</td>\n",
       "      <td>154350.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.307707</td>\n",
       "      <td>0.194028</td>\n",
       "      <td>0.068849</td>\n",
       "      <td>0.025453</td>\n",
       "      <td>0.124541</td>\n",
       "      <td>0.349846</td>\n",
       "      <td>0.359114</td>\n",
       "      <td>...</td>\n",
       "      <td>1.923476</td>\n",
       "      <td>1.625297</td>\n",
       "      <td>4.606771</td>\n",
       "      <td>4946.183888</td>\n",
       "      <td>593.337336</td>\n",
       "      <td>593.337336</td>\n",
       "      <td>0.059869</td>\n",
       "      <td>0.011887</td>\n",
       "      <td>0.909389</td>\n",
       "      <td>184.570312</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 77 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       sampling_rate  total_samples    duration  chroma_mean_0  chroma_mean_1  \\\n",
       "count          122.0     122.000000  122.000000     122.000000     122.000000   \n",
       "mean         22050.0  153183.885246    6.947115       0.256014       0.096849   \n",
       "std              0.0    9069.934393    0.411335       0.031143       0.038376   \n",
       "min          22050.0   83217.000000    3.774014       0.188875       0.004705   \n",
       "25%          22050.0  154350.000000    7.000000       0.225710       0.067572   \n",
       "50%          22050.0  154350.000000    7.000000       0.257555       0.102908   \n",
       "75%          22050.0  154350.000000    7.000000       0.283674       0.120758   \n",
       "max          22050.0  154350.000000    7.000000       0.307707       0.194028   \n",
       "\n",
       "       chroma_mean_2  chroma_mean_3  chroma_mean_4  chroma_mean_5  \\\n",
       "count     122.000000     122.000000     122.000000     122.000000   \n",
       "mean        0.018883       0.005733       0.035318       0.214208   \n",
       "std         0.012205       0.005535       0.029643       0.031075   \n",
       "min         0.003091       0.000000       0.000000       0.134658   \n",
       "25%         0.009924       0.001502       0.013512       0.193511   \n",
       "50%         0.014943       0.003994       0.028536       0.210491   \n",
       "75%         0.023548       0.008489       0.048341       0.233583   \n",
       "max         0.068849       0.025453       0.124541       0.349846   \n",
       "\n",
       "       chroma_mean_6  ...  contrast_std_4  contrast_std_5  contrast_std_6  \\\n",
       "count     122.000000  ...      122.000000      122.000000      122.000000   \n",
       "mean        0.240684  ...        1.707834        1.302730        4.088079   \n",
       "std         0.047745  ...        0.088822        0.075221        0.247915   \n",
       "min         0.130089  ...        1.508113        1.155480        3.485877   \n",
       "25%         0.212933  ...        1.656098        1.249198        3.939090   \n",
       "50%         0.241342  ...        1.705757        1.300551        4.137765   \n",
       "75%         0.259929  ...        1.769236        1.344990        4.264405   \n",
       "max         0.359114  ...        1.923476        1.625297        4.606771   \n",
       "\n",
       "       rolloff_mean  rolloff_std  rolloff_skew  zrate_mean   zrate_std  \\\n",
       "count    122.000000   122.000000    122.000000  122.000000  122.000000   \n",
       "mean    4664.787692   420.106882    420.106882    0.048327    0.008571   \n",
       "std      129.391998    68.570409     68.570409    0.003986    0.001031   \n",
       "min     4346.819300   269.640466    269.640466    0.039852    0.006078   \n",
       "25%     4568.060485   378.114109    378.114109    0.045546    0.007744   \n",
       "50%     4659.638987   414.565745    414.565745    0.048476    0.008630   \n",
       "75%     4774.292598   465.569150    465.569150    0.050921    0.009270   \n",
       "max     4946.183888   593.337336    593.337336    0.059869    0.011887   \n",
       "\n",
       "       zrate_skew       tempo  \n",
       "count  122.000000  122.000000  \n",
       "mean     0.076895  108.375900  \n",
       "std      0.216112   33.200135  \n",
       "min     -0.363726   89.102909  \n",
       "25%     -0.084345   89.102909  \n",
       "50%      0.065217   89.102909  \n",
       "75%      0.223775  123.815918  \n",
       "max      0.909389  184.570312  \n",
       "\n",
       "[8 rows x 77 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "audiofeaturesdf.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "e7beeb2e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "OK    61\n",
       "KO    61\n",
       "Name: category, dtype: int64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "audiofeaturesdf['category'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "1760e709",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Number of files')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "ax = audiofeaturesdf['category'].value_counts().plot(\n",
    "    kind='bar', figsize=(15, 7), title=\"Number of sound files per category\")\n",
    "ax.set_xlabel(\"Category\")\n",
    "ax.set_ylabel(\"Number of files\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6a1fe06",
   "metadata": {},
   "source": [
    "## 6. EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "8a7b8a71",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Let's create a sample for the EDA\n",
    "\n",
    "pctsample = 0.1\n",
    "dfsample = audiofeaturesdf.sample(frac=pctsample)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "ed3d587a",
   "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>fullfilename</th>\n",
       "      <th>dirsname</th>\n",
       "      <th>category</th>\n",
       "      <th>filename</th>\n",
       "      <th>basename</th>\n",
       "      <th>extension</th>\n",
       "      <th>sampling_rate</th>\n",
       "      <th>total_samples</th>\n",
       "      <th>duration</th>\n",
       "      <th>chroma_mean_0</th>\n",
       "      <th>...</th>\n",
       "      <th>contrast_std_4</th>\n",
       "      <th>contrast_std_5</th>\n",
       "      <th>contrast_std_6</th>\n",
       "      <th>rolloff_mean</th>\n",
       "      <th>rolloff_std</th>\n",
       "      <th>rolloff_skew</th>\n",
       "      <th>zrate_mean</th>\n",
       "      <th>zrate_std</th>\n",
       "      <th>zrate_skew</th>\n",
       "      <th>tempo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>audio_samples/KO/machine_KO_41.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_41.wav</td>\n",
       "      <td>machine_KO_41</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.260584</td>\n",
       "      <td>...</td>\n",
       "      <td>1.705195</td>\n",
       "      <td>1.341721</td>\n",
       "      <td>4.180576</td>\n",
       "      <td>4685.075975</td>\n",
       "      <td>426.956574</td>\n",
       "      <td>426.956574</td>\n",
       "      <td>0.052613</td>\n",
       "      <td>0.009588</td>\n",
       "      <td>0.212459</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>audio_samples/KO/machine_KO_39.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_39.wav</td>\n",
       "      <td>machine_KO_39</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.301744</td>\n",
       "      <td>...</td>\n",
       "      <td>1.678651</td>\n",
       "      <td>1.323780</td>\n",
       "      <td>4.160196</td>\n",
       "      <td>4666.430502</td>\n",
       "      <td>522.786233</td>\n",
       "      <td>522.786233</td>\n",
       "      <td>0.047315</td>\n",
       "      <td>0.009270</td>\n",
       "      <td>-0.002360</td>\n",
       "      <td>135.999178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>audio_samples/OK/machine_OK_61.wav</td>\n",
       "      <td>audio_samples/OK</td>\n",
       "      <td>OK</td>\n",
       "      <td>machine_OK_61.wav</td>\n",
       "      <td>machine_OK_61</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>83217</td>\n",
       "      <td>3.774014</td>\n",
       "      <td>0.225548</td>\n",
       "      <td>...</td>\n",
       "      <td>1.577671</td>\n",
       "      <td>1.219854</td>\n",
       "      <td>4.602928</td>\n",
       "      <td>4814.388270</td>\n",
       "      <td>445.426375</td>\n",
       "      <td>445.426375</td>\n",
       "      <td>0.055718</td>\n",
       "      <td>0.010076</td>\n",
       "      <td>-0.126979</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>audio_samples/OK/machine_OK_21.wav</td>\n",
       "      <td>audio_samples/OK</td>\n",
       "      <td>OK</td>\n",
       "      <td>machine_OK_21.wav</td>\n",
       "      <td>machine_OK_21</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.304858</td>\n",
       "      <td>...</td>\n",
       "      <td>1.743531</td>\n",
       "      <td>1.315048</td>\n",
       "      <td>4.147122</td>\n",
       "      <td>4563.434768</td>\n",
       "      <td>314.329218</td>\n",
       "      <td>314.329218</td>\n",
       "      <td>0.044230</td>\n",
       "      <td>0.007649</td>\n",
       "      <td>0.142765</td>\n",
       "      <td>89.102909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>audio_samples/KO/machine_KO_29.wav</td>\n",
       "      <td>audio_samples/KO</td>\n",
       "      <td>KO</td>\n",
       "      <td>machine_KO_29.wav</td>\n",
       "      <td>machine_KO_29</td>\n",
       "      <td>wav</td>\n",
       "      <td>22050</td>\n",
       "      <td>154350</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.206262</td>\n",
       "      <td>...</td>\n",
       "      <td>1.780518</td>\n",
       "      <td>1.328199</td>\n",
       "      <td>3.919138</td>\n",
       "      <td>4596.768451</td>\n",
       "      <td>525.890010</td>\n",
       "      <td>525.890010</td>\n",
       "      <td>0.049190</td>\n",
       "      <td>0.008671</td>\n",
       "      <td>-0.104756</td>\n",
       "      <td>172.265625</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 83 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                           fullfilename          dirsname category  \\\n",
       "34   audio_samples/KO/machine_KO_41.wav  audio_samples/KO       KO   \n",
       "31   audio_samples/KO/machine_KO_39.wav  audio_samples/KO       KO   \n",
       "117  audio_samples/OK/machine_OK_61.wav  audio_samples/OK       OK   \n",
       "73   audio_samples/OK/machine_OK_21.wav  audio_samples/OK       OK   \n",
       "20   audio_samples/KO/machine_KO_29.wav  audio_samples/KO       KO   \n",
       "\n",
       "              filename       basename extension  sampling_rate  total_samples  \\\n",
       "34   machine_KO_41.wav  machine_KO_41       wav          22050         154350   \n",
       "31   machine_KO_39.wav  machine_KO_39       wav          22050         154350   \n",
       "117  machine_OK_61.wav  machine_OK_61       wav          22050          83217   \n",
       "73   machine_OK_21.wav  machine_OK_21       wav          22050         154350   \n",
       "20   machine_KO_29.wav  machine_KO_29       wav          22050         154350   \n",
       "\n",
       "     duration  chroma_mean_0  ...  contrast_std_4  contrast_std_5  \\\n",
       "34   7.000000       0.260584  ...        1.705195        1.341721   \n",
       "31   7.000000       0.301744  ...        1.678651        1.323780   \n",
       "117  3.774014       0.225548  ...        1.577671        1.219854   \n",
       "73   7.000000       0.304858  ...        1.743531        1.315048   \n",
       "20   7.000000       0.206262  ...        1.780518        1.328199   \n",
       "\n",
       "     contrast_std_6  rolloff_mean  rolloff_std  rolloff_skew  zrate_mean  \\\n",
       "34         4.180576   4685.075975   426.956574    426.956574    0.052613   \n",
       "31         4.160196   4666.430502   522.786233    522.786233    0.047315   \n",
       "117        4.602928   4814.388270   445.426375    445.426375    0.055718   \n",
       "73         4.147122   4563.434768   314.329218    314.329218    0.044230   \n",
       "20         3.919138   4596.768451   525.890010    525.890010    0.049190   \n",
       "\n",
       "     zrate_std  zrate_skew       tempo  \n",
       "34    0.009588    0.212459   89.102909  \n",
       "31    0.009270   -0.002360  135.999178  \n",
       "117   0.010076   -0.126979   89.102909  \n",
       "73    0.007649    0.142765   89.102909  \n",
       "20    0.008671   -0.104756  172.265625  \n",
       "\n",
       "[5 rows x 83 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfsample.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f6edb25",
   "metadata": {},
   "source": [
    "## 7. Export features to CSV and Excel files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "d6e1e639",
   "metadata": {},
   "outputs": [],
   "source": [
    "csv_output = FEATURES_DIR + '/features.csv'\n",
    "audiofeaturesdf.to_csv(csv_output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "d2ade99a",
   "metadata": {},
   "outputs": [],
   "source": [
    "excel_output = FEATURES_DIR + '/features.xls'\n",
    "audiofeaturesdf.to_excel(excel_output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "68b3ea99",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['features.csv',\n",
       " 'features.xls',\n",
       " 'machine_KO_10_features.csv',\n",
       " 'machine_KO_11_features.csv',\n",
       " 'machine_KO_12_features.csv',\n",
       " 'machine_KO_13_features.csv',\n",
       " 'machine_KO_14_features.csv',\n",
       " 'machine_KO_15_features.csv',\n",
       " 'machine_KO_16_features.csv',\n",
       " 'machine_KO_17_features.csv',\n",
       " 'machine_KO_18_features.csv',\n",
       " 'machine_KO_19_features.csv',\n",
       " 'machine_KO_1_features.csv',\n",
       " 'machine_KO_20_features.csv',\n",
       " 'machine_KO_21_features.csv',\n",
       " 'machine_KO_22_features.csv',\n",
       " 'machine_KO_23_features.csv',\n",
       " 'machine_KO_24_features.csv',\n",
       " 'machine_KO_25_features.csv',\n",
       " 'machine_KO_26_features.csv',\n",
       " 'machine_KO_27_features.csv',\n",
       " 'machine_KO_28_features.csv',\n",
       " 'machine_KO_29_features.csv',\n",
       " 'machine_KO_2_features.csv',\n",
       " 'machine_KO_30_features.csv',\n",
       " 'machine_KO_31_features.csv',\n",
       " 'machine_KO_32_features.csv',\n",
       " 'machine_KO_33_features.csv',\n",
       " 'machine_KO_34_features.csv',\n",
       " 'machine_KO_35_features.csv',\n",
       " 'machine_KO_36_features.csv',\n",
       " 'machine_KO_37_features.csv',\n",
       " 'machine_KO_38_features.csv',\n",
       " 'machine_KO_39_features.csv',\n",
       " 'machine_KO_3_features.csv',\n",
       " 'machine_KO_40_features.csv',\n",
       " 'machine_KO_41_features.csv',\n",
       " 'machine_KO_42_features.csv',\n",
       " 'machine_KO_43_features.csv',\n",
       " 'machine_KO_44_features.csv',\n",
       " 'machine_KO_45_features.csv',\n",
       " 'machine_KO_46_features.csv',\n",
       " 'machine_KO_47_features.csv',\n",
       " 'machine_KO_48_features.csv',\n",
       " 'machine_KO_49_features.csv',\n",
       " 'machine_KO_4_features.csv',\n",
       " 'machine_KO_50_features.csv',\n",
       " 'machine_KO_51_features.csv',\n",
       " 'machine_KO_52_features.csv',\n",
       " 'machine_KO_53_features.csv',\n",
       " 'machine_KO_54_features.csv',\n",
       " 'machine_KO_55_features.csv',\n",
       " 'machine_KO_56_features.csv',\n",
       " 'machine_KO_57_features.csv',\n",
       " 'machine_KO_58_features.csv',\n",
       " 'machine_KO_59_features.csv',\n",
       " 'machine_KO_5_features.csv',\n",
       " 'machine_KO_60_features.csv',\n",
       " 'machine_KO_61_features.csv',\n",
       " 'machine_KO_6_features.csv',\n",
       " 'machine_KO_7_features.csv',\n",
       " 'machine_KO_8_features.csv',\n",
       " 'machine_KO_9_features.csv',\n",
       " 'machine_OK_10_features.csv',\n",
       " 'machine_OK_11_features.csv',\n",
       " 'machine_OK_12_features.csv',\n",
       " 'machine_OK_13_features.csv',\n",
       " 'machine_OK_14_features.csv',\n",
       " 'machine_OK_15_features.csv',\n",
       " 'machine_OK_16_features.csv',\n",
       " 'machine_OK_17_features.csv',\n",
       " 'machine_OK_18_features.csv',\n",
       " 'machine_OK_19_features.csv',\n",
       " 'machine_OK_1_features.csv',\n",
       " 'machine_OK_20_features.csv',\n",
       " 'machine_OK_21_features.csv',\n",
       " 'machine_OK_22_features.csv',\n",
       " 'machine_OK_23_features.csv',\n",
       " 'machine_OK_24_features.csv',\n",
       " 'machine_OK_25_features.csv',\n",
       " 'machine_OK_26_features.csv',\n",
       " 'machine_OK_27_features.csv',\n",
       " 'machine_OK_28_features.csv',\n",
       " 'machine_OK_29_features.csv',\n",
       " 'machine_OK_2_features.csv',\n",
       " 'machine_OK_30_features.csv',\n",
       " 'machine_OK_31_features.csv',\n",
       " 'machine_OK_32_features.csv',\n",
       " 'machine_OK_33_features.csv',\n",
       " 'machine_OK_34_features.csv',\n",
       " 'machine_OK_35_features.csv',\n",
       " 'machine_OK_36_features.csv',\n",
       " 'machine_OK_37_features.csv',\n",
       " 'machine_OK_38_features.csv',\n",
       " 'machine_OK_39_features.csv',\n",
       " 'machine_OK_3_features.csv',\n",
       " 'machine_OK_40_features.csv',\n",
       " 'machine_OK_41_features.csv',\n",
       " 'machine_OK_42_features.csv',\n",
       " 'machine_OK_43_features.csv',\n",
       " 'machine_OK_44_features.csv',\n",
       " 'machine_OK_45_features.csv',\n",
       " 'machine_OK_46_features.csv',\n",
       " 'machine_OK_47_features.csv',\n",
       " 'machine_OK_48_features.csv',\n",
       " 'machine_OK_49_features.csv',\n",
       " 'machine_OK_4_features.csv',\n",
       " 'machine_OK_50_features.csv',\n",
       " 'machine_OK_51_features.csv',\n",
       " 'machine_OK_52_features.csv',\n",
       " 'machine_OK_53_features.csv',\n",
       " 'machine_OK_54_features.csv',\n",
       " 'machine_OK_55_features.csv',\n",
       " 'machine_OK_56_features.csv',\n",
       " 'machine_OK_57_features.csv',\n",
       " 'machine_OK_58_features.csv',\n",
       " 'machine_OK_59_features.csv',\n",
       " 'machine_OK_5_features.csv',\n",
       " 'machine_OK_60_features.csv',\n",
       " 'machine_OK_61_features.csv',\n",
       " 'machine_OK_6_features.csv',\n",
       " 'machine_OK_7_features.csv',\n",
       " 'machine_OK_8_features.csv',\n",
       " 'machine_OK_9_features.csv']"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.listdir(FEATURES_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "7c618f66",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Click to download the file:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<a href='audio_features/features.csv' target='_blank'>audio_features/features.csv</a><br>"
      ],
      "text/plain": [
       "/mnt/batch/tasks/shared/LS_root/mounts/clusters/standardd13v2/code/Users/seretkow/Sound/MachineSoundAnomaly/audio_features/features.csv"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import FileLink\n",
    "\n",
    "CSVLINK = FileLink(path=csv_output)\n",
    "print(\"Click to download the file:\")\n",
    "CSVLINK"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "3f4eeea9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Click to download the file:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<a href='audio_features/features.xls' target='_blank'>audio_features/features.xls</a><br>"
      ],
      "text/plain": [
       "/mnt/batch/tasks/shared/LS_root/mounts/clusters/standardd13v2/code/Users/seretkow/Sound/MachineSoundAnomaly/audio_features/features.xls"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "XLSLINK = FileLink(path=excel_output)\n",
    "print(\"Click to download the file:\")\n",
    "XLSLINK"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3c142a84",
   "metadata": {},
   "source": [
    "> We will use AutoML for Classification in order to predict the music genre with the extracted features"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4dbb3660",
   "metadata": {},
   "source": [
    "Note: All these Python notebooks were made for demo purposes. They were not designed for production usage. Some feature can be currently in public preview. This preview version is provided without a service-level agreement. Certain features might not be supported or might have constrained capabilities. For more information, see Supplemental Terms of Use for Microsoft Azure Previews. https://azure.microsoft.com/en-us/support/legal/preview-supplemental-terms/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6887d500",
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}