Training Set Validation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training Set Validation\n",
    "Some errors were noted in the training set, so this notebook will serve to check the note values of the training set.\n",
    "\n",
    "The new training set was generated by running the newly updated `raw_to_npy.py`, which uses transposition functions from Vectorizer.py that have also been modified.\n",
    "\n",
    "The new training set is smaller because it is using the \"Major Tunes\" subset of original songs (this can be changed in `raw_to_npy.py`) however even if only common time songs are used, there is a significant improvement over the old method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Consts\n",
    "PREV_DATA = \"Common_Time_Notes\"\n",
    "CURR_DATA = \"June_Fixes_Notes\"\n",
    "PLAY_AUDIO = False #if set to true then music21 will be used to play a sample of the dataset on loading\n",
    "cutoff_percent = 0.01 # Songs will be counted as \"out of key\" if > this fraction of notes are not in the key (zero allowed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Imported 6142 songs from previous set\n",
      "Imported 3468 songs from current set\n"
     ]
    }
   ],
   "source": [
    "#Imports\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "if PLAY_AUDIO:\n",
    "    from src.Generation.Decoding import Decoding\n",
    "prev_tunes = np.load(\"Data/Vectors/{}.npy\".format(PREV_DATA))\n",
    "print(\"Imported {} songs from previous set\".format(prev_tunes.shape[0]))\n",
    "print(\"Imported {} songs from current set\".format(curr_tunes.shape[0]))\n",
    "if PLAY_AUDIO:\n",
    "    decoder = Decoding.Decoder.from_single_vector(x_train, presentation=False, time='1/32', key=\"Dmaj\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras import optimizers\n",
    "from tensorflow.keras import models\n",
    "from tensorflow.keras.layers import LeakyReLU\n",
    "from tensorflow.keras.datasets import mnist\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dicts for analysis\n",
    "note_names_to_num = {\n",
    "    'c' : 0,\n",
    "    'c#' : 1,\n",
    "    'd' : 2,\n",
    "    'd#' : 3,\n",
    "    'e' : 4,\n",
    "    'f' : 5,\n",
    "    'f#': 6,\n",
    "    'g' : 7,\n",
    "    'g#' : 8,\n",
    "    'a' : 9,\n",
    "    'a#' : 10,\n",
    "    'b' : 11,    \n",
    "}\n",
    "\n",
    "note_num_to_name = {\n",
    "    0: 'c',\n",
    "    1: 'c#',\n",
    "    2: 'd',\n",
    "    3: 'd#',\n",
    "    4: 'e',\n",
    "    5: 'f',\n",
    "    6: 'f#',\n",
    "    7: 'g',\n",
    "    8: 'g#',\n",
    "    9: 'a',\n",
    "    10: 'a#',\n",
    "    11: 'b',    \n",
    "}\n",
    "\n",
    "# Scales\n",
    "dmaj_notes = [2, 4, 6, 7, 9, 11, 1]\n",
    "dmyxo_notes = [2, 4, 6, 7, 9, 11, 0]\n",
    "d_comb = dmaj_notes + dmyxo_notes\n",
    "\n",
    "def song_in_key(single_tune_notes, key_list, verbose=False):\n",
    "    in_key = True\n",
    "    for note in single_tune_notes.flatten()%12:\n",
    "        if note not in key_list:\n",
    "            in_key = False\n",
    "            if verbose:\n",
    "                print(\"{}({}) is not in the given key\".format(note, note_num_to_name[note]))\n",
    "    return in_key\n",
    "\n",
    "def num_notes_out_of_key(single_tune_notes, key_list, verbose=False):\n",
    "    num_notes = 0\n",
    "    for note in single_tune_notes.flatten()%12:\n",
    "        if note not in key_list:\n",
    "            num_notes += 1\n",
    "            if verbose:\n",
    "                print(\"{}({}) is not in the given key\".format(note, note_num_to_name[note]))\n",
    "    return num_notes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "if PLAY_AUDIO:\n",
    "    rnd = np.random.randint(0, curr_tunes.shape[0])\n",
    "    print(\"Selected Tune#: {}\".format(rnd))\n",
    "    sample = x_train[rnd]\n",
    "    song_in_key(sample, d_comb, verbose=True)\n",
    "    decoder.play(rnd)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Count songs 'out of key'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def analyze_tunes(tune_vector, key_list, name=\"Tunes\"):\n",
    "    wrong_note_list = []\n",
    "    wrong_note_count = 0\n",
    "    cutoff_n = tune_vector.shape[0] * cutoff_percent\n",
    "    for n, tune in enumerate(tune_vector):\n",
    "        wrong_n = num_notes_out_of_key(tune, dmyxo_notes+dmaj_notes)\n",
    "        wrong_note_list.append(wrong_n)\n",
    "        if wrong_n > cutoff_n:\n",
    "            wrong_note_count += 1\n",
    "    print(\"{} of {} songs in {} exceeded the cutoff value for wrong notes. {:.2f}%\".format(wrong_note_count,tune_vector.shape[0], name, wrong_note_count/tune_vector.shape[0]*100))\n",
    "    plt.hist(wrong_note_list, range(0,600,10))\n",
    "    plt.title(\"{}: Incorrect notes, zero inclusive\".format(name))\n",
    "    plt.show()\n",
    "    plt.hist(wrong_note_list, range(1,100,1))\n",
    "    plt.title(\"{}: Incorrect notes, zero excluded\".format(name))\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2437 of 6142 songs in Previous Tunes exceeded the cutoff value for wrong notes. 39.68%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "132 of 3468 songs in Current Tunes exceeded the cutoff value for wrong notes. 3.81%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from pylab import rcParams\n",
    "rcParams['figure.figsize'] = 20, 3\n",
    "analyze_tunes(prev_tunes, d_comb, \"Previous Tunes\")\n",
    "analyze_tunes(curr_tunes, d_comb, \"Current Tunes\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}