Merge pull request #6 from spotify-song-suggestor/recommend

Moved some files for convenience, added model and encoded data. Wrote…

Pipfile

@@ -15,6 +15,8 @@ gunicorn = "*"
 plotly = "*"
 dash = "*"
 dash-bootstrap-components = "*"
+tensorflow = "*"
+sklearn = "*"
 
 [dev-packages]
 

app/dash-template/pages/predictions.py

@@ -4,22 +4,40 @@
 import dash_core_components as dcc
 import dash_html_components as html
 from dash.dependencies import Input, Output
-
+import joblib
+from tensorflow.keras.models import load_model
+import numpy as np
 # Imports from this application
 from app import app
 
+
+encodings = joblib.load(r'app\dash-template\assets\encoded_data.joblib')
+knn = joblib.load(r'app\dash-template\assets\knn.joblib')
+model = load_model(r'app\dash-template\assets\ae4')
+
+
+def recommend(index: int, n: int=5):
+    '''
+    index: index of song
+    n: number of recommendations to pull
+    returns: array of distances, array of indexes for recommended songs. Includes
+        original song.
+    '''
+    return knn.kneighbors([encodings[index]], n_neighbors=5)
+
+
+def get_songs(indeces: 'list[int]'):
+    ''' TODO
+    Uses SQLAlchemy queries to return track data from their indeces
+    '''
+    return None
+
+
 # 2 column layout. 1st column width = 4/12
 # https://dash-bootstrap-components.opensource.faculty.ai/l/components/layout
 column1 = dbc.Col(
     [
         dcc.Markdown(
-            """
-        
-            ## Predictions
-
-            Your instructions: How to use your app to get new predictions.
-
-            """
         ),
     ],
     md=4,
@@ -31,4 +49,4 @@
     ]
 )
 
-layout = dbc.Row([column1, column2])
+layout = dbc.Row([column1, column2])

model.ipynb

@@ -14,8 +14,8 @@
   },
   "orig_nbformat": 2,
   "kernelspec": {
-   "name": "python388jvsc74a57bd03dbf5d5b7d5bb3e55d0bec78b06bc8b9b99cd7236bee819b06a33ed77bfc0ab5",
-   "display_name": "Python 3.8.8 64-bit ('DS-Unit-4-Sprint-3-Deep-Learning-bsxCo5fk': pipenv)"
+   "name": "python388jvsc74a57bd035c03aa0d851cb6285394a035bf26f38eed74d8588e29c3dd9d14089f75b8b7c",
+   "display_name": "Python 3.8.8 64-bit ('NLP-MTXprD7X': pipenv)"
   },
   "metadata": {
    "interpreter": {
@@ -58,16 +58,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 78,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [],
    "source": [
-    "df_full = pd.read_csv('model_ready_data_no_dupes.csv')"
+    "df_full = pd.read_csv(r'C:\\Users\\Logan\\Desktop\\model_ready_data_no_dupes.csv')"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 79,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -76,7 +76,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 80,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -85,7 +85,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 81,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
@@ -96,7 +96,7 @@
       ]
      },
      "metadata": {},
-     "execution_count": 81
+     "execution_count": 6
     }
    ],
    "source": [
@@ -105,7 +105,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 85,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [
     {
@@ -182,7 +182,7 @@
       "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>name</th>\n      <th>duration_ms</th>\n      <th>explicit</th>\n      <th>artists</th>\n      <th>release_date</th>\n      <th>danceability</th>\n      <th>energy</th>\n      <th>key</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>popularity</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>carve</td>\n      <td>0.022576</td>\n      <td>0</td>\n      <td>[\"'uli'\"]</td>\n      <td>1922-02-22</td>\n      <td>0.645</td>\n      <td>0.4450</td>\n      <td>0.000000</td>\n      <td>0.713748</td>\n      <td>1</td>\n      <td>0.4510</td>\n      <td>0.674</td>\n      <td>0.744000</td>\n      <td>0.1510</td>\n      <td>0.1270</td>\n      <td>0.425564</td>\n      <td>0.6</td>\n      <td>0.06</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>capítulo 2.16 - banquero anarquista</td>\n      <td>0.017470</td>\n      <td>0</td>\n      <td>[\"'fernando pessoa'\"]</td>\n      <td>1922-06-01</td>\n      <td>0.695</td>\n      <td>0.2630</td>\n      <td>0.000000</td>\n      <td>0.579173</td>\n      <td>1</td>\n      <td>0.9570</td>\n      <td>0.797</td>\n      <td>0.000000</td>\n      <td>0.1480</td>\n      <td>0.6550</td>\n      <td>0.414029</td>\n      <td>0.2</td>\n      <td>0.00</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>vivo para quererte - remasterizado</td>\n      <td>0.032313</td>\n      <td>0</td>\n      <td>[\"'ignacio corsini'\"]</td>\n      <td>1922-03-21</td>\n      <td>0.434</td>\n      <td>0.1770</td>\n      <td>0.090909</td>\n      <td>0.593796</td>\n      <td>1</td>\n      <td>0.0512</td>\n      <td>0.994</td>\n      <td>0.021800</td>\n      <td>0.2120</td>\n      <td>0.4570</td>\n      <td>0.529335</td>\n      <td>1.0</td>\n      <td>0.00</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>el prisionero - remasterizado</td>\n      <td>0.031471</td>\n      <td>0</td>\n      <td>[\"'ignacio corsini'\"]</td>\n      <td>1922-03-21</td>\n      <td>0.321</td>\n      <td>0.0946</td>\n      <td>0.636364</td>\n      <td>0.490073</td>\n      <td>1</td>\n      <td>0.0504</td>\n      <td>0.995</td>\n      <td>0.918000</td>\n      <td>0.1040</td>\n      <td>0.3970</td>\n      <td>0.689907</td>\n      <td>0.6</td>\n      <td>0.00</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>lady of the evening</td>\n      <td>0.029012</td>\n      <td>0</td>\n      <td>[\"'dick haymes'\"]</td>\n      <td>1922</td>\n      <td>0.402</td>\n      <td>0.1580</td>\n      <td>0.272727</td>\n      <td>0.659263</td>\n      <td>0</td>\n      <td>0.0390</td>\n      <td>0.989</td>\n      <td>0.130000</td>\n      <td>0.3110</td>\n      <td>0.1960</td>\n      <td>0.418945</td>\n      <td>0.8</td>\n      <td>0.00</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>524206</th>\n      <td>云与海</td>\n      <td>0.045945</td>\n      <td>0</td>\n      <td>[\"'阿yueyue'\"]</td>\n      <td>2020-09-26</td>\n      <td>0.560</td>\n      <td>0.5180</td>\n      <td>0.000000</td>\n      <td>0.803491</td>\n      <td>0</td>\n      <td>0.0292</td>\n      <td>0.785</td>\n      <td>0.000000</td>\n      <td>0.0648</td>\n      <td>0.2110</td>\n      <td>0.535333</td>\n      <td>0.8</td>\n      <td>0.50</td>\n    </tr>\n    <tr>\n      <th>524207</th>\n      <td>blind</td>\n      <td>0.027270</td>\n      <td>0</td>\n      <td>[\"'role model'\"]</td>\n      <td>2020-10-21</td>\n      <td>0.765</td>\n      <td>0.6630</td>\n      <td>0.000000</td>\n      <td>0.837876</td>\n      <td>1</td>\n      <td>0.0652</td>\n      <td>0.141</td>\n      <td>0.000297</td>\n      <td>0.0924</td>\n      <td>0.6860</td>\n      <td>0.609183</td>\n      <td>0.8</td>\n      <td>0.72</td>\n    </tr>\n    <tr>\n      <th>524208</th>\n      <td>what they'll say about us</td>\n      <td>0.033374</td>\n      <td>0</td>\n      <td>[\"'finneas'\"]</td>\n      <td>2020-09-02</td>\n      <td>0.535</td>\n      <td>0.3140</td>\n      <td>0.636364</td>\n      <td>0.721626</td>\n      <td>0</td>\n      <td>0.0408</td>\n      <td>0.895</td>\n      <td>0.000150</td>\n      <td>0.0874</td>\n      <td>0.0663</td>\n      <td>0.588905</td>\n      <td>0.8</td>\n      <td>0.70</td>\n    </tr>\n    <tr>\n      <th>524209</th>\n      <td>a day at a time</td>\n      <td>0.025262</td>\n      <td>0</td>\n      <td>[\"'gentle bones'\", \"'clara benin'\"]</td>\n      <td>2021-03-05</td>\n      <td>0.696</td>\n      <td>0.6150</td>\n      <td>0.909091</td>\n      <td>0.822748</td>\n      <td>1</td>\n      <td>0.0345</td>\n      <td>0.206</td>\n      <td>0.000003</td>\n      <td>0.3050</td>\n      <td>0.4380</td>\n      <td>0.365406</td>\n      <td>0.8</td>\n      <td>0.58</td>\n    </tr>\n    <tr>\n      <th>524210</th>\n      <td>mar de emociones</td>\n      <td>0.038134</td>\n      <td>0</td>\n      <td>[\"'afrosound'\"]</td>\n      <td>2015-07-01</td>\n      <td>0.686</td>\n      <td>0.7230</td>\n      <td>0.545455</td>\n      <td>0.809670</td>\n      <td>1</td>\n      <td>0.0363</td>\n      <td>0.105</td>\n      <td>0.000000</td>\n      <td>0.2640</td>\n      <td>0.9750</td>\n      <td>0.455408</td>\n      <td>0.8</td>\n      <td>0.38</td>\n    </tr>\n  </tbody>\n</table>\n<p>524211 rows × 18 columns</p>\n</div>"
      },
      "metadata": {},
-     "execution_count": 85
+     "execution_count": 7
     }
    ],
    "source": [
@@ -614,7 +614,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 254,
+   "execution_count": 16,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "ae4 = load_model('ae4')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
    "metadata": {},
    "outputs": [
     {
@@ -625,7 +634,7 @@
       ]
      },
      "metadata": {},
-     "execution_count": 254
+     "execution_count": 17
     }
    ],
    "source": [
@@ -634,6 +643,26 @@
     "knn4.fit(encoded_data_4)"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "metadata": {},
+   "outputs": [
+    {
+     "output_type": "execute_result",
+     "data": {
+      "text/plain": [
+       "['encoded_data.joblib']"
+      ]
+     },
+     "metadata": {},
+     "execution_count": 19
+    }
+   ],
+   "source": [
+    "joblib.dump(encoded_data_4, 'encoded_data.joblib')"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 255,
@@ -1069,6 +1098,15 @@
    "source": [
     "df_full[df_full['artists'].str.contains('slipkn') == True].head(50)"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model = load_model('ae4')"
+   ]
   }
  ]
 }