pct.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f80fa9ed-65c1-46f7-8e1d-dc3b102e6671",
   "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>Unnamed: 0</th>\n",
       "      <th>track_id</th>\n",
       "      <th>artists</th>\n",
       "      <th>album_name</th>\n",
       "      <th>track_name</th>\n",
       "      <th>popularity</th>\n",
       "      <th>duration_ms</th>\n",
       "      <th>explicit</th>\n",
       "      <th>danceability</th>\n",
       "      <th>energy</th>\n",
       "      <th>...</th>\n",
       "      <th>loudness</th>\n",
       "      <th>mode</th>\n",
       "      <th>speechiness</th>\n",
       "      <th>acousticness</th>\n",
       "      <th>instrumentalness</th>\n",
       "      <th>liveness</th>\n",
       "      <th>valence</th>\n",
       "      <th>tempo</th>\n",
       "      <th>time_signature</th>\n",
       "      <th>track_genre</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>5SuOikwiRyPMVoIQDJUgSV</td>\n",
       "      <td>Gen Hoshino</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>73</td>\n",
       "      <td>230666</td>\n",
       "      <td>False</td>\n",
       "      <td>0.676</td>\n",
       "      <td>0.4610</td>\n",
       "      <td>...</td>\n",
       "      <td>-6.746</td>\n",
       "      <td>0</td>\n",
       "      <td>0.1430</td>\n",
       "      <td>0.0322</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>0.3580</td>\n",
       "      <td>0.715</td>\n",
       "      <td>87.917</td>\n",
       "      <td>4</td>\n",
       "      <td>acoustic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>4qPNDBW1i3p13qLCt0Ki3A</td>\n",
       "      <td>Ben Woodward</td>\n",
       "      <td>Ghost (Acoustic)</td>\n",
       "      <td>Ghost - Acoustic</td>\n",
       "      <td>55</td>\n",
       "      <td>149610</td>\n",
       "      <td>False</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.1660</td>\n",
       "      <td>...</td>\n",
       "      <td>-17.235</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0763</td>\n",
       "      <td>0.9240</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>0.1010</td>\n",
       "      <td>0.267</td>\n",
       "      <td>77.489</td>\n",
       "      <td>4</td>\n",
       "      <td>acoustic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1iJBSr7s7jYXzM8EGcbK5b</td>\n",
       "      <td>Ingrid Michaelson;ZAYN</td>\n",
       "      <td>To Begin Again</td>\n",
       "      <td>To Begin Again</td>\n",
       "      <td>57</td>\n",
       "      <td>210826</td>\n",
       "      <td>False</td>\n",
       "      <td>0.438</td>\n",
       "      <td>0.3590</td>\n",
       "      <td>...</td>\n",
       "      <td>-9.734</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0557</td>\n",
       "      <td>0.2100</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.1170</td>\n",
       "      <td>0.120</td>\n",
       "      <td>76.332</td>\n",
       "      <td>4</td>\n",
       "      <td>acoustic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>6lfxq3CG4xtTiEg7opyCyx</td>\n",
       "      <td>Kina Grannis</td>\n",
       "      <td>Crazy Rich Asians (Original Motion Picture Sou...</td>\n",
       "      <td>Can't Help Falling In Love</td>\n",
       "      <td>71</td>\n",
       "      <td>201933</td>\n",
       "      <td>False</td>\n",
       "      <td>0.266</td>\n",
       "      <td>0.0596</td>\n",
       "      <td>...</td>\n",
       "      <td>-18.515</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0363</td>\n",
       "      <td>0.9050</td>\n",
       "      <td>0.000071</td>\n",
       "      <td>0.1320</td>\n",
       "      <td>0.143</td>\n",
       "      <td>181.740</td>\n",
       "      <td>3</td>\n",
       "      <td>acoustic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>5vjLSffimiIP26QG5WcN2K</td>\n",
       "      <td>Chord Overstreet</td>\n",
       "      <td>Hold On</td>\n",
       "      <td>Hold On</td>\n",
       "      <td>82</td>\n",
       "      <td>198853</td>\n",
       "      <td>False</td>\n",
       "      <td>0.618</td>\n",
       "      <td>0.4430</td>\n",
       "      <td>...</td>\n",
       "      <td>-9.681</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0526</td>\n",
       "      <td>0.4690</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0829</td>\n",
       "      <td>0.167</td>\n",
       "      <td>119.949</td>\n",
       "      <td>4</td>\n",
       "      <td>acoustic</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0                track_id                 artists  \\\n",
       "0           0  5SuOikwiRyPMVoIQDJUgSV             Gen Hoshino   \n",
       "1           1  4qPNDBW1i3p13qLCt0Ki3A            Ben Woodward   \n",
       "2           2  1iJBSr7s7jYXzM8EGcbK5b  Ingrid Michaelson;ZAYN   \n",
       "3           3  6lfxq3CG4xtTiEg7opyCyx            Kina Grannis   \n",
       "4           4  5vjLSffimiIP26QG5WcN2K        Chord Overstreet   \n",
       "\n",
       "                                          album_name  \\\n",
       "0                                             Comedy   \n",
       "1                                   Ghost (Acoustic)   \n",
       "2                                     To Begin Again   \n",
       "3  Crazy Rich Asians (Original Motion Picture Sou...   \n",
       "4                                            Hold On   \n",
       "\n",
       "                   track_name  popularity  duration_ms  explicit  \\\n",
       "0                      Comedy          73       230666     False   \n",
       "1            Ghost - Acoustic          55       149610     False   \n",
       "2              To Begin Again          57       210826     False   \n",
       "3  Can't Help Falling In Love          71       201933     False   \n",
       "4                     Hold On          82       198853     False   \n",
       "\n",
       "   danceability  energy  ...  loudness  mode  speechiness  acousticness  \\\n",
       "0         0.676  0.4610  ...    -6.746     0       0.1430        0.0322   \n",
       "1         0.420  0.1660  ...   -17.235     1       0.0763        0.9240   \n",
       "2         0.438  0.3590  ...    -9.734     1       0.0557        0.2100   \n",
       "3         0.266  0.0596  ...   -18.515     1       0.0363        0.9050   \n",
       "4         0.618  0.4430  ...    -9.681     1       0.0526        0.4690   \n",
       "\n",
       "   instrumentalness  liveness  valence    tempo  time_signature  track_genre  \n",
       "0          0.000001    0.3580    0.715   87.917               4     acoustic  \n",
       "1          0.000006    0.1010    0.267   77.489               4     acoustic  \n",
       "2          0.000000    0.1170    0.120   76.332               4     acoustic  \n",
       "3          0.000071    0.1320    0.143  181.740               3     acoustic  \n",
       "4          0.000000    0.0829    0.167  119.949               4     acoustic  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from IPython.display import display\n",
    "import numpy as np\n",
    "\n",
    "spi = pd.read_csv('./dataset/dataset.csv')\n",
    "spi.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "7bbd5dee-5503-4959-95d3-b8c2f83f8927",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unnamed: 0            int64\n",
      "popularity            int64\n",
      "duration_ms           int64\n",
      "danceability        float64\n",
      "energy              float64\n",
      "key                   int64\n",
      "loudness            float64\n",
      "mode                  int64\n",
      "speechiness         float64\n",
      "acousticness        float64\n",
      "instrumentalness    float64\n",
      "liveness            float64\n",
      "valence             float64\n",
      "tempo               float64\n",
      "time_signature        int64\n",
      "dtype: object\n"
     ]
    }
   ],
   "source": [
    "spf = spi.select_dtypes(include=['int64','float64'])\n",
    "dtrain = spf.sample(frac=0.9, random_state=0)\n",
    "dvalid = spf.drop(dtrain.index)\n",
    "print(dtrain.dtypes)\n",
    "\n",
    "#display(dtrain.head(4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "bfc828c1-6962-45ad-a1fd-a93626025111",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(102600, 14)\n"
     ]
    }
   ],
   "source": [
    "max_ = dtrain.max(axis=0)\n",
    "min_ = dtrain.min(axis=0)\n",
    "dtrain = (dtrain - min_ ) / (max_ - min_)\n",
    "dvalid = (dvalid - min_) / (max_ - min_)\n",
    "X_train = dtrain.drop('Unnamed: 0', axis=1)\n",
    "X_valid = dvalid.drop('Unnamed: 0', axis=1)\n",
    "y_train = dtrain['Unnamed: 0']\n",
    "y_valid = dvalid['Unnamed: 0']\n",
    "\n",
    "print(X_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "9e8e47ad-0642-4e7b-80eb-53fce9eb4c2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "\n",
    "model = keras.Sequential([\n",
    "    layers.Dense(512, activation='relu', input_shape=[14]),\n",
    "    layers.Dense(512, activation='relu'),\n",
    "    layers.Dense(512, activation='relu'),\n",
    "    layers.Dense(1),\n",
    "])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "a6d7d546-577b-4f32-88b2-b2789b4c77e1",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(\n",
    "    optimizer='adam',\n",
    "    loss='mae',\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "53079d9a-ec57-44fa-8e17-18b646937d78",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "201/201 [==============================] - 5s 21ms/step - loss: 0.2417 - val_loss: 0.2380\n",
      "Epoch 2/20\n",
      "201/201 [==============================] - 6s 30ms/step - loss: 0.2330 - val_loss: 0.2341\n",
      "Epoch 3/20\n",
      "201/201 [==============================] - 5s 24ms/step - loss: 0.2304 - val_loss: 0.2325\n",
      "Epoch 4/20\n",
      "201/201 [==============================] - 6s 29ms/step - loss: 0.2276 - val_loss: 0.2301\n",
      "Epoch 5/20\n",
      "201/201 [==============================] - 6s 31ms/step - loss: 0.2255 - val_loss: 0.2271\n",
      "Epoch 6/20\n",
      "201/201 [==============================] - 6s 29ms/step - loss: 0.2236 - val_loss: 0.2254\n",
      "Epoch 7/20\n",
      "201/201 [==============================] - 5s 26ms/step - loss: 0.2218 - val_loss: 0.2269\n",
      "Epoch 8/20\n",
      "201/201 [==============================] - 6s 27ms/step - loss: 0.2209 - val_loss: 0.2231\n",
      "Epoch 9/20\n",
      "201/201 [==============================] - 6s 28ms/step - loss: 0.2191 - val_loss: 0.2234\n",
      "Epoch 10/20\n",
      "201/201 [==============================] - 5s 27ms/step - loss: 0.2179 - val_loss: 0.2224\n",
      "Epoch 11/20\n",
      "201/201 [==============================] - 5s 26ms/step - loss: 0.2166 - val_loss: 0.2219\n",
      "Epoch 12/20\n",
      "201/201 [==============================] - 6s 29ms/step - loss: 0.2152 - val_loss: 0.2190\n",
      "Epoch 13/20\n",
      "201/201 [==============================] - 6s 31ms/step - loss: 0.2145 - val_loss: 0.2191\n",
      "Epoch 14/20\n",
      "201/201 [==============================] - 5s 27ms/step - loss: 0.2132 - val_loss: 0.2195\n",
      "Epoch 15/20\n",
      "201/201 [==============================] - 6s 29ms/step - loss: 0.2123 - val_loss: 0.2186\n",
      "Epoch 16/20\n",
      "201/201 [==============================] - 6s 28ms/step - loss: 0.2114 - val_loss: 0.2180\n",
      "Epoch 17/20\n",
      "201/201 [==============================] - 6s 28ms/step - loss: 0.2108 - val_loss: 0.2179\n",
      "Epoch 18/20\n",
      "201/201 [==============================] - 5s 26ms/step - loss: 0.2090 - val_loss: 0.2163\n",
      "Epoch 19/20\n",
      "201/201 [==============================] - 5s 26ms/step - loss: 0.2081 - val_loss: 0.2157\n",
      "Epoch 20/20\n",
      "201/201 [==============================] - 6s 28ms/step - loss: 0.2071 - val_loss: 0.2147\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(\n",
    "    X_train, y_train,\n",
    "    validation_data=(X_valid, y_valid),\n",
    "    batch_size=512,\n",
    "    epochs=20,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "d8bd8ef3-7207-43c6-80b4-2dba315a34bd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# convert the training history to a dataframe\n",
    "history_df = pd.DataFrame(history.history)\n",
    "# use Pandas native plot method\n",
    "history_df['loss'].plot();\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "0d3ffaf7-194e-4f49-9f6f-209b79ba6a11",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "history_df['val_loss'].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "83f6b475-20a0-4212-a7db-8a35d9478d39",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Existing dir deleted\n",
      "data created\n"
     ]
    }
   ],
   "source": [
    "# Save the model to disk\n",
    "import os\n",
    "import sys\n",
    "import shutil\n",
    "\n",
    "save_dir = 'trainedmodels'\n",
    "if os.path.exists(save_dir):\n",
    "    shutil.rmtree(save_dir)\n",
    "    print('Existing dir deleted')\n",
    "    os.makedirs(save_dir)\n",
    "else:\n",
    "    pass\n",
    "    \n",
    "model.save(os.path.join(save_dir,'highqualitymodel.keras'))\n",
    "print(\"data created\")\n",
    "\n",
    "\n",
    "# Later, load the model for deployment\n",
    "#loaded_model = keras.models.load_model('path_to_saved_model')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "8c1aef7d-ef83-4401-916a-eb7a8ddc40e1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "401/401 [==============================] - 5s 14ms/step - loss: 0.2107 - val_loss: 0.2206\n",
      "Epoch 2/10\n",
      "401/401 [==============================] - 6s 14ms/step - loss: 0.2096 - val_loss: 0.2162\n",
      "Epoch 3/10\n",
      "401/401 [==============================] - 6s 15ms/step - loss: 0.2073 - val_loss: 0.2152\n",
      "Epoch 4/10\n",
      "401/401 [==============================] - 6s 16ms/step - loss: 0.2063 - val_loss: 0.2140\n",
      "Epoch 5/10\n",
      "401/401 [==============================] - 6s 16ms/step - loss: 0.2054 - val_loss: 0.2147\n",
      "Epoch 6/10\n",
      "401/401 [==============================] - 7s 17ms/step - loss: 0.2040 - val_loss: 0.2140\n",
      "Epoch 7/10\n",
      "401/401 [==============================] - 7s 17ms/step - loss: 0.2029 - val_loss: 0.2130\n",
      "Epoch 8/10\n",
      "401/401 [==============================] - 7s 16ms/step - loss: 0.2020 - val_loss: 0.2128\n",
      "Epoch 9/10\n",
      "401/401 [==============================] - 7s 16ms/step - loss: 0.2012 - val_loss: 0.2132\n",
      "Epoch 10/10\n",
      "401/401 [==============================] - 7s 16ms/step - loss: 0.1997 - val_loss: 0.2133\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(\n",
    "    X_train, y_train,\n",
    "    validation_data=(X_valid, y_valid),\n",
    "    batch_size=256,\n",
    "    epochs=10,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "028454c4-1956-48df-9137-83657c17e5a3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# convert the training history to a dataframe\n",
    "history_df = pd.DataFrame(history.history)\n",
    "# use Pandas native plot method\n",
    "history_df['loss'].plot();\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "a7a76805-e85a-477f-a544-a6a006e1a5d6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "history_df['val_loss'].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "cae9670c-1136-4941-a2f6-8ee1035c5532",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "data created\n"
     ]
    }
   ],
   "source": [
    "# Save the model to disk\n",
    "import os\n",
    "import sys\n",
    "import shutil\n",
    "\n",
    "save_dir = 'trainedmodels'\n",
    "#if os.path.exists(save_dir):\n",
    " #   shutil.rmtree(save_dir)\n",
    "  #  print('Existing dir deleted')\n",
    "    #os.makedirs(save_dir)\n",
    "#else:\n",
    " #   pass\n",
    "    \n",
    "model.save(os.path.join(save_dir,'lowqualitymodel.keras'))\n",
    "print(\"data created\")\n",
    "\n",
    "\n",
    "# Later, load the model for deployment\n",
    "#loaded_model = keras.models.load_model('path_to_saved_model')\n"
   ]

  }
 ],
 "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": "3.10.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}