Giraffe.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_output = np.array([0.72, -0.23, 1.4, 2.2, -2.3, 0.1, 0.2, 0.5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "def image_to_song(image):\n",
    "    # song         image\n",
    "    # \"solemnity\", fear\n",
    "    # \"tenderness\", amusement\n",
    "    # \"nostalgia\", awe\n",
    "    # \"calmness\", contentment\n",
    "    # \"power\", anger\n",
    "    # \"joyful activation\", excitement\n",
    "    # \"tension\", disgust\n",
    "    # \"sadness\" sadness\n",
    "    \n",
    "    # image order \n",
    "    #     labels = [\n",
    "    #         'amusement',\n",
    "    #         'anger',\n",
    "    #         'awe',\n",
    "    #         'contentment',\n",
    "    #         'disgust',\n",
    "    #         'excitement',\n",
    "    #         'fear',\n",
    "    #         'sadness'\n",
    "    #     ]\n",
    "    \n",
    "    amusement = image[:, 0]\n",
    "    anger = image[:, 1]\n",
    "    awe = image[:, 2]\n",
    "    contentment = image[:, 3]\n",
    "    disgust = image[:, 4]\n",
    "    excitement = image[:, 5]\n",
    "    fear = image[:, 6]\n",
    "    sadness = image[:, 7]\n",
    "    \n",
    "    song = [\n",
    "        fear,\n",
    "        amusement,\n",
    "        awe,\n",
    "        contentment,\n",
    "        anger,\n",
    "        excitement,\n",
    "        disgust,\n",
    "        sadness\n",
    "    ]\n",
    "    \n",
    "    return np.column_stack(song)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.2 ,  0.72,  1.4 ,  2.2 , -0.23,  0.1 , -2.3 ,  0.5 ],\n",
       "       [ 3.2 ,  3.72,  4.4 ,  5.2 ,  2.77,  3.1 ,  0.7 ,  3.5 ]])"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_songs = image_to_song(np.stack([image_output, image_output + 3]))\n",
    "test_songs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "def closest_song(song, all_songs, labels):\n",
    "    distances = np.linalg.norm(all_songs-song, axis=1)\n",
    "    return labels[np.argmin(distances)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'random song A'"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "closest_song(np.array([0, 0, 1, 2, 2, 0, -2, 0]), test_songs, ['random song A', 'giraffe'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['./Songs/Dame Da Ne.mp3',\n",
       " './Songs/National Anthem of USSR.mp3',\n",
       " './Songs/A$AP Rocky - L$D (LOVE x $EX x DREAMS).mp3',\n",
       " './Songs/BTS(방탄소년단)DOPE(쩔어).mp3',\n",
       " './Songs/Noak Hellsing - Lost.mp3',\n",
       " './Songs/Gurenge - Lisa.mp3',\n",
       " './Songs/Mac Miller - Good News.mp3',\n",
       " './Songs/Bastille - Those Nights.mp3',\n",
       " './Songs/Martin Garrix feat. Macklemore & Patrick Stump of Fall Out Boy - Summer Days.mp3',\n",
       " './Songs/Lauv - Changes.mp3',\n",
       " './Songs/ODESZA - Line Of Sight (feat. WYNNE & Mansionair).mp3',\n",
       " './Songs/Never Gonna Give You Up.mp3',\n",
       " './Songs/In The End - Linkin Park.mp3',\n",
       " './Songs/Lauv - Sims.mp3',\n",
       " './Songs/Doja Cat - Say So.mp3',\n",
       " './Songs/Lil Mosey - Blueberry Faygo (Dir. by ColeBennett).mp3',\n",
       " './Songs/Sheck Wes - Mo Bamba.mp3',\n",
       " './Songs/마마무 (MAMAMOO) - 딩가딩가 (Dingga).mp3',\n",
       " './Songs/Disturbed - The Sound Of Silence.mp3',\n",
       " './Songs/Steve Jablonsky - Arrival To Earth.mp3',\n",
       " './Songs/LeeSSang(리쌍) Pursuing The Happiness(행복을 찾아서) (Feat. Jo Hyuna of Urban Zakapa).mp3',\n",
       " \"./Songs/Richard Clayderman - Mariage d'amour.mp3\",\n",
       " './Songs/Jake Scott - Man Who Stays.mp3',\n",
       " './Songs/Pharrell Williams - Happy.mp3',\n",
       " './Songs/love ride - christian french.mp3',\n",
       " './Songs/Justin Bieber & benny blanco - Lonely.mp3',\n",
       " './Songs/DaBaby - BOP on Broadway.mp3',\n",
       " './Songs/Sam Smith - To Die For.mp3',\n",
       " './Songs/Emilee - Heaven.mp3',\n",
       " './Songs/Rich & Sad.mp3',\n",
       " './Songs/John Barrowman - I know him so well (duet with Daniel Boys).mp3',\n",
       " './Songs/Serendipity.mp3',\n",
       " './Songs/Hippie Sabotage - Devil Eyes.mp3',\n",
       " './Songs/Bruno Mars - It Will Rain.mp3',\n",
       " './Songs/Cartoon - On & On (feat. Daniel Levi).mp3',\n",
       " './Songs/Lil Nas X - HOLIDAY.mp3',\n",
       " './Songs/Dreamville - Down Bad feat. J.I.D, Bas, J. Cole, EarthGang, & Young Nudy.mp3',\n",
       " './Songs/NAV & Gunna ft. Travis Scott - Turks.mp3',\n",
       " './Songs/Mako - Smoke Filled Room.mp3']"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "songs = []\n",
    "for filename in os.listdir('./Songs'):\n",
    "    if filename.endswith('.mp3'):\n",
    "        songs.append('./Songs/' + filename)\n",
    "songs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# quick hack\n",
    "import sys\n",
    "sys.path.insert(1, './SongEmotionModel/cloudfunction')\n",
    "import main as sem\n",
    "import torch.nn.functional as F"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict_song(path):\n",
    "    image = sem.song_to_image(path)\n",
    "    out = sem.model_conv(image)\n",
    "    softmax_out = F.softmax(out)\n",
    "    return softmax_out.tolist()[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "('./Songs/Dame Da Ne.mp3',\n",
       " [0.08135166764259338,\n",
       "  0.04336877167224884,\n",
       "  0.5321084856987,\n",
       "  0.011995078064501286,\n",
       "  0.12835630774497986,\n",
       "  0.16865086555480957,\n",
       "  0.02240850031375885,\n",
       "  0.01176037173718214])"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "songs[0], predict_song(songs[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/axes/_axes.py:7592: RuntimeWarning: divide by zero encountered in log10\n",
      "  Z = 10. * np.log10(spec)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/librosa/core/audio.py:161: UserWarning: PySoundFile failed. Trying audioread instead.\n",
      "  warnings.warn('PySoundFile failed. Trying audioread instead.')\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:452: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:459: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/matplotlib/image.py:464: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "<string>:6: UserWarning: Warning: converting a masked element to nan.\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/numpy/core/_asarray.py:85: UserWarning: Warning: converting a masked element to nan.\n",
      "  return array(a, dtype, copy=False, order=order)\n",
      "/home/smxu/memories/venv/lib/python3.6/site-packages/ipykernel_launcher.py:4: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "  after removing the cwd from sys.path.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[[0.08135166764259338,\n",
       "  0.04336877167224884,\n",
       "  0.5321084856987,\n",
       "  0.011995078064501286,\n",
       "  0.12835630774497986,\n",
       "  0.16865086555480957,\n",
       "  0.02240850031375885,\n",
       "  0.01176037173718214],\n",
       " [0.06995171308517456,\n",
       "  0.029827609658241272,\n",
       "  0.35480940341949463,\n",
       "  0.009526905603706837,\n",
       "  0.07423509657382965,\n",
       "  0.38734152913093567,\n",
       "  0.06787853688001633,\n",
       "  0.0064292047172784805],\n",
       " [0.2948155403137207,\n",
       "  0.024665385484695435,\n",
       "  0.08877885341644287,\n",
       "  0.026507604867219925,\n",
       "  0.4266516864299774,\n",
       "  0.116752989590168,\n",
       "  0.01724783517420292,\n",
       "  0.004580083303153515],\n",
       " [0.0821915790438652,\n",
       "  0.025134939700365067,\n",
       "  0.2092188000679016,\n",
       "  0.018801789730787277,\n",
       "  0.2581463158130646,\n",
       "  0.3774268627166748,\n",
       "  0.019192717969417572,\n",
       "  0.009886987507343292],\n",
       " [0.16910454630851746,\n",
       "  0.03304067999124527,\n",
       "  0.2017359733581543,\n",
       "  0.009505491703748703,\n",
       "  0.4237564504146576,\n",
       "  0.14248871803283691,\n",
       "  0.010969246737658978,\n",
       "  0.009398800320923328],\n",
       " [0.043832335621118546,\n",
       "  0.08426789939403534,\n",
       "  0.35601484775543213,\n",
       "  0.015337634831666946,\n",
       "  0.10773827880620956,\n",
       "  0.3689343333244324,\n",
       "  0.014777359552681446,\n",
       "  0.009097308851778507],\n",
       " [0.20753318071365356,\n",
       "  0.05617297440767288,\n",
       "  0.15453733503818512,\n",
       "  0.011326789855957031,\n",
       "  0.33493322134017944,\n",
       "  0.16774065792560577,\n",
       "  0.062306419014930725,\n",
       "  0.005449393764138222],\n",
       " [0.11463767290115356,\n",
       "  0.05043032765388489,\n",
       "  0.2853781282901764,\n",
       "  0.010931098833680153,\n",
       "  0.2188553810119629,\n",
       "  0.2438371479511261,\n",
       "  0.06662045419216156,\n",
       "  0.009309840388596058],\n",
       " [0.15020395815372467,\n",
       "  0.03147527202963829,\n",
       "  0.339664489030838,\n",
       "  0.01133977435529232,\n",
       "  0.2697370946407318,\n",
       "  0.1605442762374878,\n",
       "  0.026808535680174828,\n",
       "  0.010226646438241005],\n",
       " [0.0664956197142601,\n",
       "  0.02334890328347683,\n",
       "  0.4576791524887085,\n",
       "  0.014103059656918049,\n",
       "  0.2460014373064041,\n",
       "  0.1466267853975296,\n",
       "  0.02456078864634037,\n",
       "  0.021184301003813744],\n",
       " [0.12391887605190277,\n",
       "  0.014485560357570648,\n",
       "  0.27153611183166504,\n",
       "  0.007450961973518133,\n",
       "  0.32919877767562866,\n",
       "  0.10156317800283432,\n",
       "  0.1444769948720932,\n",
       "  0.0073694633319973946],\n",
       " [0.12852491438388824,\n",
       "  0.041729576885700226,\n",
       "  0.31193411350250244,\n",
       "  0.007732874248176813,\n",
       "  0.2925758957862854,\n",
       "  0.16490468382835388,\n",
       "  0.03925958648324013,\n",
       "  0.013338381424546242],\n",
       " [0.2030363380908966,\n",
       "  0.04134870693087578,\n",
       "  0.2405918538570404,\n",
       "  0.016992080956697464,\n",
       "  0.30768200755119324,\n",
       "  0.11653713881969452,\n",
       "  0.061520520597696304,\n",
       "  0.0122913159430027],\n",
       " [0.34597980976104736,\n",
       "  0.045045848935842514,\n",
       "  0.17342044413089752,\n",
       "  0.01566055417060852,\n",
       "  0.2010466307401657,\n",
       "  0.18887874484062195,\n",
       "  0.026805032044649124,\n",
       "  0.003162933746352792],\n",
       " [0.2598998546600342,\n",
       "  0.030401231721043587,\n",
       "  0.27861133217811584,\n",
       "  0.01163402572274208,\n",
       "  0.28412190079689026,\n",
       "  0.07910791039466858,\n",
       "  0.045569539070129395,\n",
       "  0.010654169134795666],\n",
       " [0.12445607781410217,\n",
       "  0.04959399253129959,\n",
       "  0.2621399164199829,\n",
       "  0.01586935482919216,\n",
       "  0.19620691239833832,\n",
       "  0.31663233041763306,\n",
       "  0.02965967170894146,\n",
       "  0.005441788583993912],\n",
       " [0.11856776475906372,\n",
       "  0.05658585578203201,\n",
       "  0.27929213643074036,\n",
       "  0.017844416201114655,\n",
       "  0.15262342989444733,\n",
       "  0.326006680727005,\n",
       "  0.04376283660531044,\n",
       "  0.005316921044141054],\n",
       " [0.2422766089439392,\n",
       "  0.08124459534883499,\n",
       "  0.1297740340232849,\n",
       "  0.009599441662430763,\n",
       "  0.17196844518184662,\n",
       "  0.3092763125896454,\n",
       "  0.054818738251924515,\n",
       "  0.0010417752200737596],\n",
       " [0.2422766089439392,\n",
       "  0.08124459534883499,\n",
       "  0.1297740340232849,\n",
       "  0.009599441662430763,\n",
       "  0.17196844518184662,\n",
       "  0.3092763125896454,\n",
       "  0.054818738251924515,\n",
       "  0.0010417752200737596],\n",
       " [0.23935195803642273,\n",
       "  0.024867527186870575,\n",
       "  0.1333516389131546,\n",
       "  0.02374122105538845,\n",
       "  0.3627525568008423,\n",
       "  0.16993646323680878,\n",
       "  0.039602357894182205,\n",
       "  0.006396242883056402],\n",
       " [0.2696361839771271,\n",
       "  0.02918395958840847,\n",
       "  0.3898049294948578,\n",
       "  0.024891141802072525,\n",
       "  0.18136180937290192,\n",
       "  0.0893528088927269,\n",
       "  0.011268933303654194,\n",
       "  0.0045003206469118595],\n",
       " [0.17579616606235504,\n",
       "  0.037334565073251724,\n",
       "  0.19853641092777252,\n",
       "  0.019810829311609268,\n",
       "  0.25549009442329407,\n",
       "  0.2655597925186157,\n",
       "  0.041387178003787994,\n",
       "  0.006084998603910208],\n",
       " [0.3762146830558777,\n",
       "  0.033867962658405304,\n",
       "  0.23411568999290466,\n",
       "  0.008839746937155724,\n",
       "  0.16850750148296356,\n",
       "  0.16350902616977692,\n",
       "  0.012918943539261818,\n",
       "  0.0020264179911464453],\n",
       " [0.13174326717853546,\n",
       "  0.016691260039806366,\n",
       "  0.3560076057910919,\n",
       "  0.009330655448138714,\n",
       "  0.16694442927837372,\n",
       "  0.2843572497367859,\n",
       "  0.032774340361356735,\n",
       "  0.002151145599782467],\n",
       " [0.06044897809624672,\n",
       "  0.04439333826303482,\n",
       "  0.21431942284107208,\n",
       "  0.01715661957859993,\n",
       "  0.555677056312561,\n",
       "  0.08335337042808533,\n",
       "  0.017264775931835175,\n",
       "  0.00738657359033823],\n",
       " [0.2694146931171417,\n",
       "  0.0323178693652153,\n",
       "  0.1619511842727661,\n",
       "  0.014754677191376686,\n",
       "  0.3212586045265198,\n",
       "  0.14240795373916626,\n",
       "  0.056158095598220825,\n",
       "  0.0017368217231705785],\n",
       " [0.08205293118953705,\n",
       "  0.059244923293590546,\n",
       "  0.3175131678581238,\n",
       "  0.019954869523644447,\n",
       "  0.060045674443244934,\n",
       "  0.4234420955181122,\n",
       "  0.03673431649804115,\n",
       "  0.0010120561346411705],\n",
       " [0.08205293118953705,\n",
       "  0.059244923293590546,\n",
       "  0.3175131678581238,\n",
       "  0.019954869523644447,\n",
       "  0.060045674443244934,\n",
       "  0.4234420955181122,\n",
       "  0.03673431649804115,\n",
       "  0.0010120561346411705],\n",
       " [0.06869431585073471,\n",
       "  0.076640285551548,\n",
       "  0.2249065339565277,\n",
       "  0.0061828638426959515,\n",
       "  0.07396315783262253,\n",
       "  0.5276547074317932,\n",
       "  0.015949953347444534,\n",
       "  0.006008228287100792],\n",
       " [0.24335752427577972,\n",
       "  0.03561367467045784,\n",
       "  0.21690845489501953,\n",
       "  0.013581291772425175,\n",
       "  0.31676623225212097,\n",
       "  0.12671656906604767,\n",
       "  0.04103429615497589,\n",
       "  0.006022006738930941],\n",
       " [0.07297270745038986,\n",
       "  0.05813523009419441,\n",
       "  0.3422943353652954,\n",
       "  0.015058325603604317,\n",
       "  0.08598954230546951,\n",
       "  0.35580703616142273,\n",
       "  0.05563724413514137,\n",
       "  0.014105561189353466],\n",
       " [0.10817328840494156,\n",
       "  0.03569386526942253,\n",
       "  0.2827393710613251,\n",
       "  0.01508351694792509,\n",
       "  0.30636465549468994,\n",
       "  0.20906959474086761,\n",
       "  0.026928238570690155,\n",
       "  0.015947502106428146],\n",
       " [0.11954496800899506,\n",
       "  0.09474773705005646,\n",
       "  0.3760313391685486,\n",
       "  0.014183945953845978,\n",
       "  0.15684083104133606,\n",
       "  0.21157504618167877,\n",
       "  0.017797747626900673,\n",
       "  0.009278328157961369],\n",
       " [0.1925726979970932,\n",
       "  0.05048063397407532,\n",
       "  0.2917718291282654,\n",
       "  0.008770991116762161,\n",
       "  0.15951764583587646,\n",
       "  0.25369036197662354,\n",
       "  0.03931985795497894,\n",
       "  0.0038760367315262556],\n",
       " [0.2622518539428711,\n",
       "  0.016759905964136124,\n",
       "  0.12451136857271194,\n",
       "  0.030348705127835274,\n",
       "  0.5023719668388367,\n",
       "  0.04117973521351814,\n",
       "  0.010696745477616787,\n",
       "  0.01187978871166706],\n",
       " [0.11630937457084656,\n",
       "  0.03330976516008377,\n",
       "  0.16399982571601868,\n",
       "  0.015291064977645874,\n",
       "  0.18165339529514313,\n",
       "  0.37694069743156433,\n",
       "  0.10771610587835312,\n",
       "  0.004779765848070383],\n",
       " [0.02903117425739765,\n",
       "  0.009628918021917343,\n",
       "  0.1350066214799881,\n",
       "  0.004613203462213278,\n",
       "  0.7447620034217834,\n",
       "  0.051452986896038055,\n",
       "  0.015118327923119068,\n",
       "  0.010386807844042778],\n",
       " [0.06309404969215393,\n",
       "  0.024084964767098427,\n",
       "  0.2565677762031555,\n",
       "  0.01023037824779749,\n",
       "  0.1560894250869751,\n",
       "  0.43666142225265503,\n",
       "  0.041784368455410004,\n",
       "  0.0114876888692379],\n",
       " [0.13906419277191162,\n",
       "  0.010033893398940563,\n",
       "  0.17245425283908844,\n",
       "  0.011467301286756992,\n",
       "  0.2982407510280609,\n",
       "  0.3214247524738312,\n",
       "  0.0403037965297699,\n",
       "  0.007011079229414463]]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "song_vectors = []\n",
    "for song in songs:\n",
    "    song_vectors.append(predict_song(song))\n",
    "song_vectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.08135167, 0.04336877, 0.53210849, 0.01199508, 0.12835631,\n",
       "        0.16865087, 0.0224085 , 0.01176037],\n",
       "       [0.06995171, 0.02982761, 0.3548094 , 0.00952691, 0.0742351 ,\n",
       "        0.38734153, 0.06787854, 0.0064292 ],\n",
       "       [0.29481554, 0.02466539, 0.08877885, 0.0265076 , 0.42665169,\n",
       "        0.11675299, 0.01724784, 0.00458008],\n",
       "       [0.08219158, 0.02513494, 0.2092188 , 0.01880179, 0.25814632,\n",
       "        0.37742686, 0.01919272, 0.00988699],\n",
       "       [0.16910455, 0.03304068, 0.20173597, 0.00950549, 0.42375645,\n",
       "        0.14248872, 0.01096925, 0.0093988 ],\n",
       "       [0.04383234, 0.0842679 , 0.35601485, 0.01533763, 0.10773828,\n",
       "        0.36893433, 0.01477736, 0.00909731],\n",
       "       [0.20753318, 0.05617297, 0.15453734, 0.01132679, 0.33493322,\n",
       "        0.16774066, 0.06230642, 0.00544939],\n",
       "       [0.11463767, 0.05043033, 0.28537813, 0.0109311 , 0.21885538,\n",
       "        0.24383715, 0.06662045, 0.00930984],\n",
       "       [0.15020396, 0.03147527, 0.33966449, 0.01133977, 0.26973709,\n",
       "        0.16054428, 0.02680854, 0.01022665],\n",
       "       [0.06649562, 0.0233489 , 0.45767915, 0.01410306, 0.24600144,\n",
       "        0.14662679, 0.02456079, 0.0211843 ],\n",
       "       [0.12391888, 0.01448556, 0.27153611, 0.00745096, 0.32919878,\n",
       "        0.10156318, 0.14447699, 0.00736946],\n",
       "       [0.12852491, 0.04172958, 0.31193411, 0.00773287, 0.2925759 ,\n",
       "        0.16490468, 0.03925959, 0.01333838],\n",
       "       [0.20303634, 0.04134871, 0.24059185, 0.01699208, 0.30768201,\n",
       "        0.11653714, 0.06152052, 0.01229132],\n",
       "       [0.34597981, 0.04504585, 0.17342044, 0.01566055, 0.20104663,\n",
       "        0.18887874, 0.02680503, 0.00316293],\n",
       "       [0.25989985, 0.03040123, 0.27861133, 0.01163403, 0.2841219 ,\n",
       "        0.07910791, 0.04556954, 0.01065417],\n",
       "       [0.12445608, 0.04959399, 0.26213992, 0.01586935, 0.19620691,\n",
       "        0.31663233, 0.02965967, 0.00544179],\n",
       "       [0.11856776, 0.05658586, 0.27929214, 0.01784442, 0.15262343,\n",
       "        0.32600668, 0.04376284, 0.00531692],\n",
       "       [0.24227661, 0.0812446 , 0.12977403, 0.00959944, 0.17196845,\n",
       "        0.30927631, 0.05481874, 0.00104178],\n",
       "       [0.24227661, 0.0812446 , 0.12977403, 0.00959944, 0.17196845,\n",
       "        0.30927631, 0.05481874, 0.00104178],\n",
       "       [0.23935196, 0.02486753, 0.13335164, 0.02374122, 0.36275256,\n",
       "        0.16993646, 0.03960236, 0.00639624],\n",
       "       [0.26963618, 0.02918396, 0.38980493, 0.02489114, 0.18136181,\n",
       "        0.08935281, 0.01126893, 0.00450032],\n",
       "       [0.17579617, 0.03733457, 0.19853641, 0.01981083, 0.25549009,\n",
       "        0.26555979, 0.04138718, 0.006085  ],\n",
       "       [0.37621468, 0.03386796, 0.23411569, 0.00883975, 0.1685075 ,\n",
       "        0.16350903, 0.01291894, 0.00202642],\n",
       "       [0.13174327, 0.01669126, 0.35600761, 0.00933066, 0.16694443,\n",
       "        0.28435725, 0.03277434, 0.00215115],\n",
       "       [0.06044898, 0.04439334, 0.21431942, 0.01715662, 0.55567706,\n",
       "        0.08335337, 0.01726478, 0.00738657],\n",
       "       [0.26941469, 0.03231787, 0.16195118, 0.01475468, 0.3212586 ,\n",
       "        0.14240795, 0.0561581 , 0.00173682],\n",
       "       [0.08205293, 0.05924492, 0.31751317, 0.01995487, 0.06004567,\n",
       "        0.4234421 , 0.03673432, 0.00101206],\n",
       "       [0.08205293, 0.05924492, 0.31751317, 0.01995487, 0.06004567,\n",
       "        0.4234421 , 0.03673432, 0.00101206],\n",
       "       [0.06869432, 0.07664029, 0.22490653, 0.00618286, 0.07396316,\n",
       "        0.52765471, 0.01594995, 0.00600823],\n",
       "       [0.24335752, 0.03561367, 0.21690845, 0.01358129, 0.31676623,\n",
       "        0.12671657, 0.0410343 , 0.00602201],\n",
       "       [0.07297271, 0.05813523, 0.34229434, 0.01505833, 0.08598954,\n",
       "        0.35580704, 0.05563724, 0.01410556],\n",
       "       [0.10817329, 0.03569387, 0.28273937, 0.01508352, 0.30636466,\n",
       "        0.20906959, 0.02692824, 0.0159475 ],\n",
       "       [0.11954497, 0.09474774, 0.37603134, 0.01418395, 0.15684083,\n",
       "        0.21157505, 0.01779775, 0.00927833],\n",
       "       [0.1925727 , 0.05048063, 0.29177183, 0.00877099, 0.15951765,\n",
       "        0.25369036, 0.03931986, 0.00387604],\n",
       "       [0.26225185, 0.01675991, 0.12451137, 0.03034871, 0.50237197,\n",
       "        0.04117974, 0.01069675, 0.01187979],\n",
       "       [0.11630937, 0.03330977, 0.16399983, 0.01529106, 0.1816534 ,\n",
       "        0.3769407 , 0.10771611, 0.00477977],\n",
       "       [0.02903117, 0.00962892, 0.13500662, 0.0046132 , 0.744762  ,\n",
       "        0.05145299, 0.01511833, 0.01038681],\n",
       "       [0.06309405, 0.02408496, 0.25656778, 0.01023038, 0.15608943,\n",
       "        0.43666142, 0.04178437, 0.01148769],\n",
       "       [0.13906419, 0.01003389, 0.17245425, 0.0114673 , 0.29824075,\n",
       "        0.32142475, 0.0403038 , 0.00701108]])"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "song_array = np.array(song_vectors)\n",
    "song_array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'./Songs/Dreamville - Down Bad feat. J.I.D, Bas, J. Cole, EarthGang, & Young Nudy.mp3'"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "closest_song(np.array([0, 0, 1, 2, 2, 0, -2, 0]), song_array, songs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('./songs.npy', song_array)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Dame Da Ne',\n",
       " 'National Anthem of USSR',\n",
       " 'A$AP Rocky - L$D (LOVE x $EX x DREAMS)',\n",
       " 'BTS(방탄소년단)DOPE(쩔어)',\n",
       " 'Noak Hellsing - Lost',\n",
       " 'Gurenge - Lisa',\n",
       " 'Mac Miller - Good News',\n",
       " 'Bastille - Those Nights',\n",
       " 'Martin Garrix feat. Macklemore & Patrick Stump of Fall Out Boy - Summer Days',\n",
       " 'Lauv - Changes',\n",
       " 'ODESZA - Line Of Sight (feat. WYNNE & Mansionair)',\n",
       " 'Never Gonna Give You U',\n",
       " 'In The End - Linkin Park',\n",
       " 'Lauv - Sims',\n",
       " 'Doja Cat - Say So',\n",
       " 'Lil Mosey - Blueberry Faygo (Dir. by ColeBennett)',\n",
       " 'heck Wes - Mo Bamba',\n",
       " '마마무 (MAMAMOO) - 딩가딩가 (Dingga)',\n",
       " 'Disturbed - The Sound Of Silence',\n",
       " 'teve Jablonsky - Arrival To Earth',\n",
       " 'LeeSSang(리쌍) Pursuing The Happiness(행복을 찾아서) (Feat. Jo Hyuna of Urban Zakapa)',\n",
       " \"Richard Clayderman - Mariage d'amour\",\n",
       " 'Jake Scott - Man Who Stays',\n",
       " 'Pharrell Williams - Happy',\n",
       " 'love ride - christian french',\n",
       " 'Justin Bieber & benny blanco - Lonely',\n",
       " 'DaBaby - BOP on Broadway',\n",
       " 'am Smith - To Die For',\n",
       " 'Emilee - Heaven',\n",
       " 'Rich & Sad',\n",
       " 'John Barrowman - I know him so well (duet with Daniel Boys)',\n",
       " 'erendipity',\n",
       " 'Hippie Sabotage - Devil Eyes',\n",
       " 'Bruno Mars - It Will Rain',\n",
       " 'Cartoon - On & On (feat. Daniel Levi)',\n",
       " 'Lil Nas X - HOLIDAY',\n",
       " 'Dreamville - Down Bad feat. J.I.D, Bas, J. Cole, EarthGang, & Young Nudy',\n",
       " 'NAV & Gunna ft. Travis Scott - Turks',\n",
       " 'Mako - Smoke Filled Roo']"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[song.strip('./Songs/').strip('.mp3') for song in songs]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict(image):\n",
    "    # predict a song given a\n",
    "    song = image_to_song(image)\n",
    "    return closest_song(song, song_array, songs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'./Songs/Dame Da Ne.mp3'"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predict(np.array([[0, 0, 1, 2, 2, 0, -2, 0]]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "memories",
   "language": "python",
   "name": "memories"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}