Hugging Face's logo

Spaces:
nlovoldegar
/
karaoke_generator
Runtime error

App Files Community
nlovoldegar commited on
Commit
dc354d0
•
1 Parent(s): eeee601

Upload 3 files

Browse files
Files changed (3) hide show
  1. finalapp.ipynb +168 -0
  2. preprocessed_data.ipynb +2062 -0
  3. requirements.txt +4 -0
finalapp.ipynb ADDED
@@ -0,0 +1,168 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cells": [
3
+ {
4
+ "cell_type": "code",
5
+ "execution_count": null,
6
+ "id": "ae69f2ba",
7
+ "metadata": {},
8
+ "outputs": [],
9
+ "source": [
10
+ "import streamlit as st\n",
11
+ "st.set_page_config(page_title=\"Karaoke Playlist Generator\", layout=\"wide\")\n",
12
+ "st.markdown \"Plan your performance\"\n",
13
+ "import pandas as pd\n",
14
+ "from sklearn.neighbors import NearestNeighbors\n",
15
+ "import plotly.express as px\n",
16
+ "import streamlit.components.v1 as components\n",
17
+ "\n",
18
+ "@st.cache(allow_output_mutation=True)\n",
19
+ "def load_data():\n",
20
+ " df = pd.read_csv(\"data/filtered_track_df.csv\")\n",
21
+ " df['genres'] = df.genres.apply(lambda x: [i[1:-1] for i in str(x)[1:-1].split(\", \")])\n",
22
+ " exploded_track_df = df.explode(\"genres\")\n",
23
+ " return exploded_track_df\n",
24
+ "\n",
25
+ "genre_names = ['Dance Pop', 'Electronic', 'Electropop', 'Hip Hop', 'Jazz', 'K-pop', 'Latin', 'Pop', 'Pop Rap', 'R&B', 'Rock']\n",
26
+ "audio_feats = [\"acousticness\", \"danceability\", \"energy\", \"instrumentalness\", \"valence\", \"tempo\"]\n",
27
+ "\n",
28
+ "exploded_track_df = load_data()\n",
29
+ "\n",
30
+ "def n_neighbors_uri_audio(genre, start_year, end_year, test_feat):\n",
31
+ " genre = genre.lower()\n",
32
+ " genre_data = exploded_track_df[(exploded_track_df[\"genres\"]==genre) & (exploded_track_df[\"release_year\"]>=start_year) & (exploded_track_df[\"release_year\"]<=end_year)]\n",
33
+ " genre_data = genre_data.sort_values(by='popularity', ascending=False)[:500]\n",
34
+ "\n",
35
+ " neigh = NearestNeighbors()\n",
36
+ " neigh.fit(genre_data[audio_feats].to_numpy())\n",
37
+ "\n",
38
+ " n_neighbors = neigh.kneighbors([test_feat], n_neighbors=len(genre_data), return_distance=False)[0]\n",
39
+ "\n",
40
+ " uris = genre_data.iloc[n_neighbors][\"uri\"].tolist()\n",
41
+ " audios = genre_data.iloc[n_neighbors][audio_feats].to_numpy()\n",
42
+ " return uris, audios\n",
43
+ "\n",
44
+ "\n",
45
+ "title = \"Karaoke Performance Geneartp\"\n",
46
+ "st.title(title)\n",
47
+ "\n",
48
+ "st.write(\"Customize your performance based on genre and several key audio features!\")\n",
49
+ "st.markdown(\"##\")\n",
50
+ "\n",
51
+ "with st.container():\n",
52
+ " col1, col2,col3,col4 = st.columns((2,0.5,0.5,0.5))\n",
53
+ " with col3:\n",
54
+ " st.markdown(\"***Choose your genre:***\")\n",
55
+ " genre = st.radio(\n",
56
+ " \"\",\n",
57
+ " genre_names, index=genre_names.index(\"Pop\"))\n",
58
+ " with col1:\n",
59
+ " st.markdown(\"***Choose features to customize:***\")\n",
60
+ " start_year, end_year = st.slider(\n",
61
+ " 'Select the year range',\n",
62
+ " 1990, 2019, (2015, 2019)\n",
63
+ " )\n",
64
+ " acousticness = st.slider(\n",
65
+ " 'Acousticness',\n",
66
+ " 0.0, 1.0, 0.5)\n",
67
+ " danceability = st.slider(\n",
68
+ " 'Danceability',\n",
69
+ " 0.0, 1.0, 0.5)\n",
70
+ " energy = st.slider(\n",
71
+ " 'Energy',\n",
72
+ " 0.0, 1.0, 0.5)\n",
73
+ " instrumentalness = st.slider(\n",
74
+ " 'Instrumentalness',\n",
75
+ " 0.0, 1.0, 0.0)\n",
76
+ " valence = st.slider(\n",
77
+ " 'Valence',\n",
78
+ " 0.0, 1.0, 0.45)\n",
79
+ " tempo = st.slider(\n",
80
+ " 'Tempo',\n",
81
+ " 0.0, 244.0, 118.0)\n",
82
+ "\n",
83
+ "tracks_per_page = 6\n",
84
+ "test_feat = [acousticness, danceability, energy, instrumentalness, valence, tempo]\n",
85
+ "uris, audios = n_neighbors_uri_audio(genre, start_year, end_year, test_feat)\n",
86
+ "\n",
87
+ "tracks = []\n",
88
+ "for uri in uris:\n",
89
+ " track = \"\"\"<iframe src=\"https://open.spotify.com/embed/track/{}\" width=\"260\" height=\"380\" frameborder=\"0\" allowtransparency=\"true\" allow=\"encrypted-media\"></iframe>\"\"\".format(uri)\n",
90
+ " tracks.append(track)\n",
91
+ "\n",
92
+ "if 'previous_inputs' not in st.session_state:\n",
93
+ " st.session_state['previous_inputs'] = [genre, start_year, end_year] + test_feat\n",
94
+ "\n",
95
+ "current_inputs = [genre, start_year, end_year] + test_feat\n",
96
+ "if current_inputs != st.session_state['previous_inputs']:\n",
97
+ " if 'start_track_i' in st.session_state:\n",
98
+ " st.session_state['start_track_i'] = 0\n",
99
+ " st.session_state['previous_inputs'] = current_inputs\n",
100
+ "\n",
101
+ "if 'start_track_i' not in st.session_state:\n",
102
+ " st.session_state['start_track_i'] = 0\n",
103
+ "\n",
104
+ "with st.container():\n",
105
+ " col1, col2, col3 = st.columns([2,1,2])\n",
106
+ " if st.button(\"Recommend More Songs\"):\n",
107
+ " if st.session_state['start_track_i'] < len(tracks):\n",
108
+ " st.session_state['start_track_i'] += tracks_per_page\n",
109
+ "\n",
110
+ " current_tracks = tracks[st.session_state['start_track_i']: st.session_state['start_track_i'] + tracks_per_page]\n",
111
+ " current_audios = audios[st.session_state['start_track_i']: st.session_state['start_track_i'] + tracks_per_page]\n",
112
+ " if st.session_state['start_track_i'] < len(tracks):\n",
113
+ " for i, (track, audio) in enumerate(zip(current_tracks, current_audios)):\n",
114
+ " if i%2==0:\n",
115
+ " with col1:\n",
116
+ " components.html(\n",
117
+ " track,\n",
118
+ " height=400,\n",
119
+ " )\n",
120
+ " with st.expander(\"See more details\"):\n",
121
+ " df = pd.DataFrame(dict(\n",
122
+ " r=audio[:5],\n",
123
+ " theta=audio_feats[:5]))\n",
124
+ " fig = px.line_polar(df, r='r', theta='theta', line_close=True)\n",
125
+ " fig.update_layout(height=400, width=340)\n",
126
+ " st.plotly_chart(fig)\n",
127
+ " \n",
128
+ " else:\n",
129
+ " with col3:\n",
130
+ " components.html(\n",
131
+ " track,\n",
132
+ " height=400,\n",
133
+ " )\n",
134
+ " with st.expander(\"See more details\"):\n",
135
+ " df = pd.DataFrame(dict(\n",
136
+ " r=audio[:5],\n",
137
+ " theta=audio_feats[:5]))\n",
138
+ " fig = px.line_polar(df, r='r', theta='theta', line_close=True)\n",
139
+ " fig.update_layout(height=400, width=340)\n",
140
+ " st.plotly_chart(fig)\n",
141
+ "\n",
142
+ " else:\n",
143
+ " st.write(\"No songs left to recommend\")"
144
+ ]
145
+ }
146
+ ],
147
+ "metadata": {
148
+ "kernelspec": {
149
+ "display_name": "Python 3 (ipykernel)",
150
+ "language": "python",
151
+ "name": "python3"
152
+ },
153
+ "language_info": {
154
+ "codemirror_mode": {
155
+ "name": "ipython",
156
+ "version": 3
157
+ },
158
+ "file_extension": ".py",
159
+ "mimetype": "text/x-python",
160
+ "name": "python",
161
+ "nbconvert_exporter": "python",
162
+ "pygments_lexer": "ipython3",
163
+ "version": "3.8.13"
164
+ }
165
+ },
166
+ "nbformat": 4,
167
+ "nbformat_minor": 5
168
+ }
preprocessed_data.ipynb ADDED
@@ -0,0 +1,2062 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "nbformat": 4,
3
+ "nbformat_minor": 0,
4
+ "metadata": {
5
+ "colab": {
6
+ "provenance": []
7
+ },
8
+ "kernelspec": {
9
+ "name": "python3",
10
+ "display_name": "Python 3"
11
+ },
12
+ "language_info": {
13
+ "name": "python"
14
+ }
15
+ },
16
+ "cells": [
17
+ {
18
+ "cell_type": "code",
19
+ "execution_count": 1,
20
+ "metadata": {
21
+ "id": "FplIdD623HAp"
22
+ },
23
+ "outputs": [],
24
+ "source": [
25
+ "import pandas as pd"
26
+ ]
27
+ },
28
+ {
29
+ "cell_type": "code",
30
+ "source": [
31
+ "from google.colab import files\n",
32
+ "files.upload()"
33
+ ],
34
+ "metadata": {
35
+ "colab": {
36
+ "base_uri": "https://localhost:8080/",
37
+ "height": 91
38
+ },
39
+ "id": "1Q3uH4fM5kEC",
40
+ "outputId": "66ca70c8-4e9f-4da4-ef3e-1e50c96023d6"
41
+ },
42
+ "execution_count": 3,
43
+ "outputs": [
44
+ {
45
+ "output_type": "display_data",
46
+ "data": {
47
+ "text/plain": [
48
+ "<IPython.core.display.HTML object>"
49
+ ],
50
+ "text/html": [
51
+ "\n",
52
+ " <input type=\"file\" id=\"files-1551534a-a390-4a9d-84fd-7429e378414c\" name=\"files[]\" multiple disabled\n",
53
+ " style=\"border:none\" />\n",
54
+ " <output id=\"result-1551534a-a390-4a9d-84fd-7429e378414c\">\n",
55
+ " Upload widget is only available when the cell has been executed in the\n",
56
+ " current browser session. Please rerun this cell to enable.\n",
57
+ " </output>\n",
58
+ " <script>// Copyright 2017 Google LLC\n",
59
+ "//\n",
60
+ "// Licensed under the Apache License, Version 2.0 (the \"License\");\n",
61
+ "// you may not use this file except in compliance with the License.\n",
62
+ "// You may obtain a copy of the License at\n",
63
+ "//\n",
64
+ "// http://www.apache.org/licenses/LICENSE-2.0\n",
65
+ "//\n",
66
+ "// Unless required by applicable law or agreed to in writing, software\n",
67
+ "// distributed under the License is distributed on an \"AS IS\" BASIS,\n",
68
+ "// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
69
+ "// See the License for the specific language governing permissions and\n",
70
+ "// limitations under the License.\n",
71
+ "\n",
72
+ "/**\n",
73
+ " * @fileoverview Helpers for google.colab Python module.\n",
74
+ " */\n",
75
+ "(function(scope) {\n",
76
+ "function span(text, styleAttributes = {}) {\n",
77
+ " const element = document.createElement('span');\n",
78
+ " element.textContent = text;\n",
79
+ " for (const key of Object.keys(styleAttributes)) {\n",
80
+ " element.style[key] = styleAttributes[key];\n",
81
+ " }\n",
82
+ " return element;\n",
83
+ "}\n",
84
+ "\n",
85
+ "// Max number of bytes which will be uploaded at a time.\n",
86
+ "const MAX_PAYLOAD_SIZE = 100 * 1024;\n",
87
+ "\n",
88
+ "function _uploadFiles(inputId, outputId) {\n",
89
+ " const steps = uploadFilesStep(inputId, outputId);\n",
90
+ " const outputElement = document.getElementById(outputId);\n",
91
+ " // Cache steps on the outputElement to make it available for the next call\n",
92
+ " // to uploadFilesContinue from Python.\n",
93
+ " outputElement.steps = steps;\n",
94
+ "\n",
95
+ " return _uploadFilesContinue(outputId);\n",
96
+ "}\n",
97
+ "\n",
98
+ "// This is roughly an async generator (not supported in the browser yet),\n",
99
+ "// where there are multiple asynchronous steps and the Python side is going\n",
100
+ "// to poll for completion of each step.\n",
101
+ "// This uses a Promise to block the python side on completion of each step,\n",
102
+ "// then passes the result of the previous step as the input to the next step.\n",
103
+ "function _uploadFilesContinue(outputId) {\n",
104
+ " const outputElement = document.getElementById(outputId);\n",
105
+ " const steps = outputElement.steps;\n",
106
+ "\n",
107
+ " const next = steps.next(outputElement.lastPromiseValue);\n",
108
+ " return Promise.resolve(next.value.promise).then((value) => {\n",
109
+ " // Cache the last promise value to make it available to the next\n",
110
+ " // step of the generator.\n",
111
+ " outputElement.lastPromiseValue = value;\n",
112
+ " return next.value.response;\n",
113
+ " });\n",
114
+ "}\n",
115
+ "\n",
116
+ "/**\n",
117
+ " * Generator function which is called between each async step of the upload\n",
118
+ " * process.\n",
119
+ " * @param {string} inputId Element ID of the input file picker element.\n",
120
+ " * @param {string} outputId Element ID of the output display.\n",
121
+ " * @return {!Iterable<!Object>} Iterable of next steps.\n",
122
+ " */\n",
123
+ "function* uploadFilesStep(inputId, outputId) {\n",
124
+ " const inputElement = document.getElementById(inputId);\n",
125
+ " inputElement.disabled = false;\n",
126
+ "\n",
127
+ " const outputElement = document.getElementById(outputId);\n",
128
+ " outputElement.innerHTML = '';\n",
129
+ "\n",
130
+ " const pickedPromise = new Promise((resolve) => {\n",
131
+ " inputElement.addEventListener('change', (e) => {\n",
132
+ " resolve(e.target.files);\n",
133
+ " });\n",
134
+ " });\n",
135
+ "\n",
136
+ " const cancel = document.createElement('button');\n",
137
+ " inputElement.parentElement.appendChild(cancel);\n",
138
+ " cancel.textContent = 'Cancel upload';\n",
139
+ " const cancelPromise = new Promise((resolve) => {\n",
140
+ " cancel.onclick = () => {\n",
141
+ " resolve(null);\n",
142
+ " };\n",
143
+ " });\n",
144
+ "\n",
145
+ " // Wait for the user to pick the files.\n",
146
+ " const files = yield {\n",
147
+ " promise: Promise.race([pickedPromise, cancelPromise]),\n",
148
+ " response: {\n",
149
+ " action: 'starting',\n",
150
+ " }\n",
151
+ " };\n",
152
+ "\n",
153
+ " cancel.remove();\n",
154
+ "\n",
155
+ " // Disable the input element since further picks are not allowed.\n",
156
+ " inputElement.disabled = true;\n",
157
+ "\n",
158
+ " if (!files) {\n",
159
+ " return {\n",
160
+ " response: {\n",
161
+ " action: 'complete',\n",
162
+ " }\n",
163
+ " };\n",
164
+ " }\n",
165
+ "\n",
166
+ " for (const file of files) {\n",
167
+ " const li = document.createElement('li');\n",
168
+ " li.append(span(file.name, {fontWeight: 'bold'}));\n",
169
+ " li.append(span(\n",
170
+ " `(${file.type || 'n/a'}) - ${file.size} bytes, ` +\n",
171
+ " `last modified: ${\n",
172
+ " file.lastModifiedDate ? file.lastModifiedDate.toLocaleDateString() :\n",
173
+ " 'n/a'} - `));\n",
174
+ " const percent = span('0% done');\n",
175
+ " li.appendChild(percent);\n",
176
+ "\n",
177
+ " outputElement.appendChild(li);\n",
178
+ "\n",
179
+ " const fileDataPromise = new Promise((resolve) => {\n",
180
+ " const reader = new FileReader();\n",
181
+ " reader.onload = (e) => {\n",
182
+ " resolve(e.target.result);\n",
183
+ " };\n",
184
+ " reader.readAsArrayBuffer(file);\n",
185
+ " });\n",
186
+ " // Wait for the data to be ready.\n",
187
+ " let fileData = yield {\n",
188
+ " promise: fileDataPromise,\n",
189
+ " response: {\n",
190
+ " action: 'continue',\n",
191
+ " }\n",
192
+ " };\n",
193
+ "\n",
194
+ " // Use a chunked sending to avoid message size limits. See b/62115660.\n",
195
+ " let position = 0;\n",
196
+ " do {\n",
197
+ " const length = Math.min(fileData.byteLength - position, MAX_PAYLOAD_SIZE);\n",
198
+ " const chunk = new Uint8Array(fileData, position, length);\n",
199
+ " position += length;\n",
200
+ "\n",
201
+ " const base64 = btoa(String.fromCharCode.apply(null, chunk));\n",
202
+ " yield {\n",
203
+ " response: {\n",
204
+ " action: 'append',\n",
205
+ " file: file.name,\n",
206
+ " data: base64,\n",
207
+ " },\n",
208
+ " };\n",
209
+ "\n",
210
+ " let percentDone = fileData.byteLength === 0 ?\n",
211
+ " 100 :\n",
212
+ " Math.round((position / fileData.byteLength) * 100);\n",
213
+ " percent.textContent = `${percentDone}% done`;\n",
214
+ "\n",
215
+ " } while (position < fileData.byteLength);\n",
216
+ " }\n",
217
+ "\n",
218
+ " // All done.\n",
219
+ " yield {\n",
220
+ " response: {\n",
221
+ " action: 'complete',\n",
222
+ " }\n",
223
+ " };\n",
224
+ "}\n",
225
+ "\n",
226
+ "scope.google = scope.google || {};\n",
227
+ "scope.google.colab = scope.google.colab || {};\n",
228
+ "scope.google.colab._files = {\n",
229
+ " _uploadFiles,\n",
230
+ " _uploadFilesContinue,\n",
231
+ "};\n",
232
+ "})(self);\n",
233
+ "</script> "
234
+ ]
235
+ },
236
+ "metadata": {}
237
+ },
238
+ {
239
+ "output_type": "stream",
240
+ "name": "stdout",
241
+ "text": [
242
+ "Saving kaggle.json to kaggle.json\n"
243
+ ]
244
+ },
245
+ {
246
+ "output_type": "execute_result",
247
+ "data": {
248
+ "text/plain": [
249
+ "{'kaggle.json': b'{\"username\":\"nicolelovoldegar\",\"key\":\"6f3b82ca462a786dec38b73cb86062fc\"}'}"
250
+ ]
251
+ },
252
+ "metadata": {},
253
+ "execution_count": 3
254
+ }
255
+ ]
256
+ },
257
+ {
258
+ "cell_type": "code",
259
+ "source": [
260
+ "!mkdir ~/.kaggle/\n",
261
+ "!cp kaggle.json ~/.kaggle/"
262
+ ],
263
+ "metadata": {
264
+ "id": "XIfkGtcN5veF"
265
+ },
266
+ "execution_count": 5,
267
+ "outputs": []
268
+ },
269
+ {
270
+ "cell_type": "code",
271
+ "source": [
272
+ "!chmod 600 ~/.kaggle/kaggle.json"
273
+ ],
274
+ "metadata": {
275
+ "id": "Lv7BUFzs5so1"
276
+ },
277
+ "execution_count": 6,
278
+ "outputs": []
279
+ },
280
+ {
281
+ "cell_type": "code",
282
+ "source": [
283
+ "!kaggle datasets download -d saurabhshahane/spotgen-music-dataset"
284
+ ],
285
+ "metadata": {
286
+ "colab": {
287
+ "base_uri": "https://localhost:8080/"
288
+ },
289
+ "id": "yW3gI56q5x9B",
290
+ "outputId": "20280982-d21a-494a-915b-a554c5f76573"
291
+ },
292
+ "execution_count": 7,
293
+ "outputs": [
294
+ {
295
+ "output_type": "stream",
296
+ "name": "stdout",
297
+ "text": [
298
+ "Downloading spotgen-music-dataset.zip to /content\n",
299
+ "100% 273M/274M [00:18<00:00, 23.0MB/s]\n",
300
+ "100% 274M/274M [00:18<00:00, 15.9MB/s]\n"
301
+ ]
302
+ }
303
+ ]
304
+ },
305
+ {
306
+ "cell_type": "code",
307
+ "source": [
308
+ "!pip install patool"
309
+ ],
310
+ "metadata": {
311
+ "colab": {
312
+ "base_uri": "https://localhost:8080/"
313
+ },
314
+ "id": "eIsVsVF_57MO",
315
+ "outputId": "4a4de23c-f8b2-48ae-c7da-2483b79b0358"
316
+ },
317
+ "execution_count": 8,
318
+ "outputs": [
319
+ {
320
+ "output_type": "stream",
321
+ "name": "stdout",
322
+ "text": [
323
+ "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
324
+ "Collecting patool\n",
325
+ " Downloading patool-1.12-py2.py3-none-any.whl (77 kB)\n",
326
+ "\u001b[K |████████████████████████████████| 77 kB 3.0 MB/s \n",
327
+ "\u001b[?25hInstalling collected packages: patool\n",
328
+ "Successfully installed patool-1.12\n"
329
+ ]
330
+ }
331
+ ]
332
+ },
333
+ {
334
+ "cell_type": "code",
335
+ "source": [
336
+ "import patoolib"
337
+ ],
338
+ "metadata": {
339
+ "id": "y5kv8NsL59W6"
340
+ },
341
+ "execution_count": 9,
342
+ "outputs": []
343
+ },
344
+ {
345
+ "cell_type": "code",
346
+ "source": [
347
+ "patoolib.extract_archive('spotgen-music-dataset.zip')"
348
+ ],
349
+ "metadata": {
350
+ "colab": {
351
+ "base_uri": "https://localhost:8080/",
352
+ "height": 88
353
+ },
354
+ "id": "CCM-qXKe5_Km",
355
+ "outputId": "f2c3ac9f-c1ab-4997-ed11-d318f2967d3c"
356
+ },
357
+ "execution_count": 10,
358
+ "outputs": [
359
+ {
360
+ "output_type": "stream",
361
+ "name": "stdout",
362
+ "text": [
363
+ "patool: Extracting spotgen-music-dataset.zip ...\n",
364
+ "patool: running /usr/bin/7z x -o./Unpack_51crkv6y -- spotgen-music-dataset.zip\n",
365
+ "patool: ... spotgen-music-dataset.zip extracted to `SpotGenTrack'.\n"
366
+ ]
367
+ },
368
+ {
369
+ "output_type": "execute_result",
370
+ "data": {
371
+ "text/plain": [
372
+ "'SpotGenTrack'"
373
+ ],
374
+ "application/vnd.google.colaboratory.intrinsic+json": {
375
+ "type": "string"
376
+ }
377
+ },
378
+ "metadata": {},
379
+ "execution_count": 10
380
+ }
381
+ ]
382
+ },
383
+ {
384
+ "cell_type": "code",
385
+ "source": [
386
+ "data_dir = \"SpotGenTrack/Data Sources/\"\n",
387
+ "albums_data = pd.read_csv(data_dir + \"spotify_albums.csv\")\n",
388
+ "artists_data = pd.read_csv(data_dir + \"spotify_artists.csv\")\n",
389
+ "tracks_data = pd.read_csv(data_dir + \"spotify_tracks.csv\")"
390
+ ],
391
+ "metadata": {
392
+ "id": "q5JCr7WS5XiT"
393
+ },
394
+ "execution_count": 11,
395
+ "outputs": []
396
+ },
397
+ {
398
+ "cell_type": "code",
399
+ "source": [
400
+ "display(albums_data.head())\n",
401
+ "albums_data.columns"
402
+ ],
403
+ "metadata": {
404
+ "colab": {
405
+ "base_uri": "https://localhost:8080/",
406
+ "height": 582
407
+ },
408
+ "id": "rrQEjWwK6W2g",
409
+ "outputId": "984e20a8-720e-475c-b97d-23aebb518c4e"
410
+ },
411
+ "execution_count": 12,
412
+ "outputs": [
413
+ {
414
+ "output_type": "display_data",
415
+ "data": {
416
+ "text/plain": [
417
+ " Unnamed: 0 album_type artist_id \\\n",
418
+ "0 0 single 3DiDSECUqqY1AuBP8qtaIa \n",
419
+ "1 1 album 6s1pCNXcbdtQJlsnM1hRIA \n",
420
+ "2 2 single 5YjfNaHq05WrwldRe1QSBc \n",
421
+ "3 3 single 2G9Vc16JCpnZmK4uGH46Fa \n",
422
+ "4 4 single 2dwM9OcE4c3Ph1UBINSodx \n",
423
+ "\n",
424
+ " available_markets \\\n",
425
+ "0 ['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH... \n",
426
+ "1 ['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH... \n",
427
+ "2 ['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH... \n",
428
+ "3 ['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH... \n",
429
+ "4 ['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH... \n",
430
+ "\n",
431
+ " external_urls \\\n",
432
+ "0 {'spotify': 'https://open.spotify.com/album/1g... \n",
433
+ "1 {'spotify': 'https://open.spotify.com/album/4K... \n",
434
+ "2 {'spotify': 'https://open.spotify.com/album/7n... \n",
435
+ "3 {'spotify': 'https://open.spotify.com/album/6p... \n",
436
+ "4 {'spotify': 'https://open.spotify.com/album/1X... \n",
437
+ "\n",
438
+ " href id \\\n",
439
+ "0 https://api.spotify.com/v1/albums/1gAM7M4rBwEb... 1gAM7M4rBwEbSPeAQR2nx1 \n",
440
+ "1 https://api.spotify.com/v1/albums/4KfJZV7WfolY... 4KfJZV7WfolYlxBzOTo66s \n",
441
+ "2 https://api.spotify.com/v1/albums/7nLYY7uAVUb5... 7nLYY7uAVUb57kpd7tZxnS \n",
442
+ "3 https://api.spotify.com/v1/albums/6p20Rt4x2Qn5... 6p20Rt4x2Qn5mUMRi1s6pj \n",
443
+ "4 https://api.spotify.com/v1/albums/1XeoOqC1q7U2... 1XeoOqC1q7U2iyLEQJ64cu \n",
444
+ "\n",
445
+ " images \\\n",
446
+ "0 [{'height': 640, 'url': 'https://i.scdn.co/ima... \n",
447
+ "1 [{'height': 640, 'url': 'https://i.scdn.co/ima... \n",
448
+ "2 [{'height': 640, 'url': 'https://i.scdn.co/ima... \n",
449
+ "3 [{'height': 640, 'url': 'https://i.scdn.co/ima... \n",
450
+ "4 [{'height': 640, 'url': 'https://i.scdn.co/ima... \n",
451
+ "\n",
452
+ " name release_date \\\n",
453
+ "0 If I Ain't Got You EP 2019-02-08 \n",
454
+ "1 Shostakovich Symphony No.5 - Four Romances on ... 2019-03-01 \n",
455
+ "2 Take My Bass 2019-03-14 \n",
456
+ "3 Hypnotizing (Are U) 2016-11-16 \n",
457
+ "4 Sunshine 2018-07-20 \n",
458
+ "\n",
459
+ " release_date_precision total_tracks track_id \\\n",
460
+ "0 day 6 2iejTMy9XZ8Gaae0aQ2yl0 \n",
461
+ "1 day 8 1WQfghEjszJJ4H8MAWrQ2C \n",
462
+ "2 day 1 3jJKj4QTK3v18ZSwpk7AcV \n",
463
+ "3 day 1 1xGtDafUZbHyYC3Xarcbrj \n",
464
+ "4 day 1 0gWtsXvXOzAT6FtM3ur8in \n",
465
+ "\n",
466
+ " track_name_prev uri type \n",
467
+ "0 track_32 spotify:album:1gAM7M4rBwEbSPeAQR2nx1 album \n",
468
+ "1 track_11 spotify:album:4KfJZV7WfolYlxBzOTo66s album \n",
469
+ "2 track_15 spotify:album:7nLYY7uAVUb57kpd7tZxnS album \n",
470
+ "3 track_46 spotify:album:6p20Rt4x2Qn5mUMRi1s6pj album \n",
471
+ "4 track_10 spotify:album:1XeoOqC1q7U2iyLEQJ64cu album "
472
+ ],
473
+ "text/html": [
474
+ "\n",
475
+ " <div id=\"df-eabe1c5a-d6e1-4b7b-83a0-6578c7fa299e\">\n",
476
+ " <div class=\"colab-df-container\">\n",
477
+ " <div>\n",
478
+ "<style scoped>\n",
479
+ " .dataframe tbody tr th:only-of-type {\n",
480
+ " vertical-align: middle;\n",
481
+ " }\n",
482
+ "\n",
483
+ " .dataframe tbody tr th {\n",
484
+ " vertical-align: top;\n",
485
+ " }\n",
486
+ "\n",
487
+ " .dataframe thead th {\n",
488
+ " text-align: right;\n",
489
+ " }\n",
490
+ "</style>\n",
491
+ "<table border=\"1\" class=\"dataframe\">\n",
492
+ " <thead>\n",
493
+ " <tr style=\"text-align: right;\">\n",
494
+ " <th></th>\n",
495
+ " <th>Unnamed: 0</th>\n",
496
+ " <th>album_type</th>\n",
497
+ " <th>artist_id</th>\n",
498
+ " <th>available_markets</th>\n",
499
+ " <th>external_urls</th>\n",
500
+ " <th>href</th>\n",
501
+ " <th>id</th>\n",
502
+ " <th>images</th>\n",
503
+ " <th>name</th>\n",
504
+ " <th>release_date</th>\n",
505
+ " <th>release_date_precision</th>\n",
506
+ " <th>total_tracks</th>\n",
507
+ " <th>track_id</th>\n",
508
+ " <th>track_name_prev</th>\n",
509
+ " <th>uri</th>\n",
510
+ " <th>type</th>\n",
511
+ " </tr>\n",
512
+ " </thead>\n",
513
+ " <tbody>\n",
514
+ " <tr>\n",
515
+ " <th>0</th>\n",
516
+ " <td>0</td>\n",
517
+ " <td>single</td>\n",
518
+ " <td>3DiDSECUqqY1AuBP8qtaIa</td>\n",
519
+ " <td>['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH...</td>\n",
520
+ " <td>{'spotify': 'https://open.spotify.com/album/1g...</td>\n",
521
+ " <td>https://api.spotify.com/v1/albums/1gAM7M4rBwEb...</td>\n",
522
+ " <td>1gAM7M4rBwEbSPeAQR2nx1</td>\n",
523
+ " <td>[{'height': 640, 'url': 'https://i.scdn.co/ima...</td>\n",
524
+ " <td>If I Ain't Got You EP</td>\n",
525
+ " <td>2019-02-08</td>\n",
526
+ " <td>day</td>\n",
527
+ " <td>6</td>\n",
528
+ " <td>2iejTMy9XZ8Gaae0aQ2yl0</td>\n",
529
+ " <td>track_32</td>\n",
530
+ " <td>spotify:album:1gAM7M4rBwEbSPeAQR2nx1</td>\n",
531
+ " <td>album</td>\n",
532
+ " </tr>\n",
533
+ " <tr>\n",
534
+ " <th>1</th>\n",
535
+ " <td>1</td>\n",
536
+ " <td>album</td>\n",
537
+ " <td>6s1pCNXcbdtQJlsnM1hRIA</td>\n",
538
+ " <td>['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH...</td>\n",
539
+ " <td>{'spotify': 'https://open.spotify.com/album/4K...</td>\n",
540
+ " <td>https://api.spotify.com/v1/albums/4KfJZV7WfolY...</td>\n",
541
+ " <td>4KfJZV7WfolYlxBzOTo66s</td>\n",
542
+ " <td>[{'height': 640, 'url': 'https://i.scdn.co/ima...</td>\n",
543
+ " <td>Shostakovich Symphony No.5 - Four Romances on ...</td>\n",
544
+ " <td>2019-03-01</td>\n",
545
+ " <td>day</td>\n",
546
+ " <td>8</td>\n",
547
+ " <td>1WQfghEjszJJ4H8MAWrQ2C</td>\n",
548
+ " <td>track_11</td>\n",
549
+ " <td>spotify:album:4KfJZV7WfolYlxBzOTo66s</td>\n",
550
+ " <td>album</td>\n",
551
+ " </tr>\n",
552
+ " <tr>\n",
553
+ " <th>2</th>\n",
554
+ " <td>2</td>\n",
555
+ " <td>single</td>\n",
556
+ " <td>5YjfNaHq05WrwldRe1QSBc</td>\n",
557
+ " <td>['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH...</td>\n",
558
+ " <td>{'spotify': 'https://open.spotify.com/album/7n...</td>\n",
559
+ " <td>https://api.spotify.com/v1/albums/7nLYY7uAVUb5...</td>\n",
560
+ " <td>7nLYY7uAVUb57kpd7tZxnS</td>\n",
561
+ " <td>[{'height': 640, 'url': 'https://i.scdn.co/ima...</td>\n",
562
+ " <td>Take My Bass</td>\n",
563
+ " <td>2019-03-14</td>\n",
564
+ " <td>day</td>\n",
565
+ " <td>1</td>\n",
566
+ " <td>3jJKj4QTK3v18ZSwpk7AcV</td>\n",
567
+ " <td>track_15</td>\n",
568
+ " <td>spotify:album:7nLYY7uAVUb57kpd7tZxnS</td>\n",
569
+ " <td>album</td>\n",
570
+ " </tr>\n",
571
+ " <tr>\n",
572
+ " <th>3</th>\n",
573
+ " <td>3</td>\n",
574
+ " <td>single</td>\n",
575
+ " <td>2G9Vc16JCpnZmK4uGH46Fa</td>\n",
576
+ " <td>['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH...</td>\n",
577
+ " <td>{'spotify': 'https://open.spotify.com/album/6p...</td>\n",
578
+ " <td>https://api.spotify.com/v1/albums/6p20Rt4x2Qn5...</td>\n",
579
+ " <td>6p20Rt4x2Qn5mUMRi1s6pj</td>\n",
580
+ " <td>[{'height': 640, 'url': 'https://i.scdn.co/ima...</td>\n",
581
+ " <td>Hypnotizing (Are U)</td>\n",
582
+ " <td>2016-11-16</td>\n",
583
+ " <td>day</td>\n",
584
+ " <td>1</td>\n",
585
+ " <td>1xGtDafUZbHyYC3Xarcbrj</td>\n",
586
+ " <td>track_46</td>\n",
587
+ " <td>spotify:album:6p20Rt4x2Qn5mUMRi1s6pj</td>\n",
588
+ " <td>album</td>\n",
589
+ " </tr>\n",
590
+ " <tr>\n",
591
+ " <th>4</th>\n",
592
+ " <td>4</td>\n",
593
+ " <td>single</td>\n",
594
+ " <td>2dwM9OcE4c3Ph1UBINSodx</td>\n",
595
+ " <td>['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH...</td>\n",
596
+ " <td>{'spotify': 'https://open.spotify.com/album/1X...</td>\n",
597
+ " <td>https://api.spotify.com/v1/albums/1XeoOqC1q7U2...</td>\n",
598
+ " <td>1XeoOqC1q7U2iyLEQJ64cu</td>\n",
599
+ " <td>[{'height': 640, 'url': 'https://i.scdn.co/ima...</td>\n",
600
+ " <td>Sunshine</td>\n",
601
+ " <td>2018-07-20</td>\n",
602
+ " <td>day</td>\n",
603
+ " <td>1</td>\n",
604
+ " <td>0gWtsXvXOzAT6FtM3ur8in</td>\n",
605
+ " <td>track_10</td>\n",
606
+ " <td>spotify:album:1XeoOqC1q7U2iyLEQJ64cu</td>\n",
607
+ " <td>album</td>\n",
608
+ " </tr>\n",
609
+ " </tbody>\n",
610
+ "</table>\n",
611
+ "</div>\n",
612
+ " <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-eabe1c5a-d6e1-4b7b-83a0-6578c7fa299e')\"\n",
613
+ " title=\"Convert this dataframe to an interactive table.\"\n",
614
+ " style=\"display:none;\">\n",
615
+ " \n",
616
+ " <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
617
+ " width=\"24px\">\n",
618
+ " <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
619
+ " <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
620
+ " </svg>\n",
621
+ " </button>\n",
622
+ " \n",
623
+ " <style>\n",
624
+ " .colab-df-container {\n",
625
+ " display:flex;\n",
626
+ " flex-wrap:wrap;\n",
627
+ " gap: 12px;\n",
628
+ " }\n",
629
+ "\n",
630
+ " .colab-df-convert {\n",
631
+ " background-color: #E8F0FE;\n",
632
+ " border: none;\n",
633
+ " border-radius: 50%;\n",
634
+ " cursor: pointer;\n",
635
+ " display: none;\n",
636
+ " fill: #1967D2;\n",
637
+ " height: 32px;\n",
638
+ " padding: 0 0 0 0;\n",
639
+ " width: 32px;\n",
640
+ " }\n",
641
+ "\n",
642
+ " .colab-df-convert:hover {\n",
643
+ " background-color: #E2EBFA;\n",
644
+ " box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
645
+ " fill: #174EA6;\n",
646
+ " }\n",
647
+ "\n",
648
+ " [theme=dark] .colab-df-convert {\n",
649
+ " background-color: #3B4455;\n",
650
+ " fill: #D2E3FC;\n",
651
+ " }\n",
652
+ "\n",
653
+ " [theme=dark] .colab-df-convert:hover {\n",
654
+ " background-color: #434B5C;\n",
655
+ " box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
656
+ " filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
657
+ " fill: #FFFFFF;\n",
658
+ " }\n",
659
+ " </style>\n",
660
+ "\n",
661
+ " <script>\n",
662
+ " const buttonEl =\n",
663
+ " document.querySelector('#df-eabe1c5a-d6e1-4b7b-83a0-6578c7fa299e button.colab-df-convert');\n",
664
+ " buttonEl.style.display =\n",
665
+ " google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
666
+ "\n",
667
+ " async function convertToInteractive(key) {\n",
668
+ " const element = document.querySelector('#df-eabe1c5a-d6e1-4b7b-83a0-6578c7fa299e');\n",
669
+ " const dataTable =\n",
670
+ " await google.colab.kernel.invokeFunction('convertToInteractive',\n",
671
+ " [key], {});\n",
672
+ " if (!dataTable) return;\n",
673
+ "\n",
674
+ " const docLinkHtml = 'Like what you see? Visit the ' +\n",
675
+ " '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
676
+ " + ' to learn more about interactive tables.';\n",
677
+ " element.innerHTML = '';\n",
678
+ " dataTable['output_type'] = 'display_data';\n",
679
+ " await google.colab.output.renderOutput(dataTable, element);\n",
680
+ " const docLink = document.createElement('div');\n",
681
+ " docLink.innerHTML = docLinkHtml;\n",
682
+ " element.appendChild(docLink);\n",
683
+ " }\n",
684
+ " </script>\n",
685
+ " </div>\n",
686
+ " </div>\n",
687
+ " "
688
+ ]
689
+ },
690
+ "metadata": {}
691
+ },
692
+ {
693
+ "output_type": "execute_result",
694
+ "data": {
695
+ "text/plain": [
696
+ "Index(['Unnamed: 0', 'album_type', 'artist_id', 'available_markets',\n",
697
+ " 'external_urls', 'href', 'id', 'images', 'name', 'release_date',\n",
698
+ " 'release_date_precision', 'total_tracks', 'track_id', 'track_name_prev',\n",
699
+ " 'uri', 'type'],\n",
700
+ " dtype='object')"
701
+ ]
702
+ },
703
+ "metadata": {},
704
+ "execution_count": 12
705
+ }
706
+ ]
707
+ },
708
+ {
709
+ "cell_type": "code",
710
+ "source": [
711
+ "display(artists_data.head())\n",
712
+ "artists_data.columns"
713
+ ],
714
+ "metadata": {
715
+ "colab": {
716
+ "base_uri": "https://localhost:8080/",
717
+ "height": 258
718
+ },
719
+ "id": "tAHuaO606Yjd",
720
+ "outputId": "e9feccbe-990f-444a-a16d-b00f9d0cceea"
721
+ },
722
+ "execution_count": 13,
723
+ "outputs": [
724
+ {
725
+ "output_type": "display_data",
726
+ "data": {
727
+ "text/plain": [
728
+ " Unnamed: 0 artist_popularity followers \\\n",
729
+ "0 0 44 23230 \n",
730
+ "1 1 22 313 \n",
731
+ "2 2 26 1596 \n",
732
+ "3 3 31 149 \n",
733
+ "4 4 21 11 \n",
734
+ "\n",
735
+ " genres id \\\n",
736
+ "0 ['sertanejo', 'sertanejo pop', 'sertanejo trad... 4mGnpjhqgx4RUdsIJiURdo \n",
737
+ "1 [] 1dLnVku4VQUOLswwDFvRc9 \n",
738
+ "2 ['danish pop rock'] 6YVY310fjfUzKi8hiqR7iK \n",
739
+ "3 ['uk alternative pop'] 2VElyouiCfoYPDJluzwJwK \n",
740
+ "4 ['french baroque'] 4agVy03qW8juSysCTUOuDI \n",
741
+ "\n",
742
+ " name track_id track_name_prev type \n",
743
+ "0 Juliano Cezar 0wmDmAILuW9e2aRttkl4aC track_9 artist \n",
744
+ "1 The Grenadines 4wqwj0gA8qPZKLl5WVqXml track_30 artist \n",
745
+ "2 Gangway 1bFqWDbvHmZe2f4Nf9qaD8 track_38 artist \n",
746
+ "3 FADES 3MFSUBAidPzRBbIS7BDj1S track_34 artist \n",
747
+ "4 Jean-Pierre Guignon 2r3q57FhxdsCyYr0kuDq4b track_26 artist "
748
+ ],
749
+ "text/html": [
750
+ "\n",
751
+ " <div id=\"df-d959edf2-79c3-4736-8a38-326b6c0d0f2d\">\n",
752
+ " <div class=\"colab-df-container\">\n",
753
+ " <div>\n",
754
+ "<style scoped>\n",
755
+ " .dataframe tbody tr th:only-of-type {\n",
756
+ " vertical-align: middle;\n",
757
+ " }\n",
758
+ "\n",
759
+ " .dataframe tbody tr th {\n",
760
+ " vertical-align: top;\n",
761
+ " }\n",
762
+ "\n",
763
+ " .dataframe thead th {\n",
764
+ " text-align: right;\n",
765
+ " }\n",
766
+ "</style>\n",
767
+ "<table border=\"1\" class=\"dataframe\">\n",
768
+ " <thead>\n",
769
+ " <tr style=\"text-align: right;\">\n",
770
+ " <th></th>\n",
771
+ " <th>Unnamed: 0</th>\n",
772
+ " <th>artist_popularity</th>\n",
773
+ " <th>followers</th>\n",
774
+ " <th>genres</th>\n",
775
+ " <th>id</th>\n",
776
+ " <th>name</th>\n",
777
+ " <th>track_id</th>\n",
778
+ " <th>track_name_prev</th>\n",
779
+ " <th>type</th>\n",
780
+ " </tr>\n",
781
+ " </thead>\n",
782
+ " <tbody>\n",
783
+ " <tr>\n",
784
+ " <th>0</th>\n",
785
+ " <td>0</td>\n",
786
+ " <td>44</td>\n",
787
+ " <td>23230</td>\n",
788
+ " <td>['sertanejo', 'sertanejo pop', 'sertanejo trad...</td>\n",
789
+ " <td>4mGnpjhqgx4RUdsIJiURdo</td>\n",
790
+ " <td>Juliano Cezar</td>\n",
791
+ " <td>0wmDmAILuW9e2aRttkl4aC</td>\n",
792
+ " <td>track_9</td>\n",
793
+ " <td>artist</td>\n",
794
+ " </tr>\n",
795
+ " <tr>\n",
796
+ " <th>1</th>\n",
797
+ " <td>1</td>\n",
798
+ " <td>22</td>\n",
799
+ " <td>313</td>\n",
800
+ " <td>[]</td>\n",
801
+ " <td>1dLnVku4VQUOLswwDFvRc9</td>\n",
802
+ " <td>The Grenadines</td>\n",
803
+ " <td>4wqwj0gA8qPZKLl5WVqXml</td>\n",
804
+ " <td>track_30</td>\n",
805
+ " <td>artist</td>\n",
806
+ " </tr>\n",
807
+ " <tr>\n",
808
+ " <th>2</th>\n",
809
+ " <td>2</td>\n",
810
+ " <td>26</td>\n",
811
+ " <td>1596</td>\n",
812
+ " <td>['danish pop rock']</td>\n",
813
+ " <td>6YVY310fjfUzKi8hiqR7iK</td>\n",
814
+ " <td>Gangway</td>\n",
815
+ " <td>1bFqWDbvHmZe2f4Nf9qaD8</td>\n",
816
+ " <td>track_38</td>\n",
817
+ " <td>artist</td>\n",
818
+ " </tr>\n",
819
+ " <tr>\n",
820
+ " <th>3</th>\n",
821
+ " <td>3</td>\n",
822
+ " <td>31</td>\n",
823
+ " <td>149</td>\n",
824
+ " <td>['uk alternative pop']</td>\n",
825
+ " <td>2VElyouiCfoYPDJluzwJwK</td>\n",
826
+ " <td>FADES</td>\n",
827
+ " <td>3MFSUBAidPzRBbIS7BDj1S</td>\n",
828
+ " <td>track_34</td>\n",
829
+ " <td>artist</td>\n",
830
+ " </tr>\n",
831
+ " <tr>\n",
832
+ " <th>4</th>\n",
833
+ " <td>4</td>\n",
834
+ " <td>21</td>\n",
835
+ " <td>11</td>\n",
836
+ " <td>['french baroque']</td>\n",
837
+ " <td>4agVy03qW8juSysCTUOuDI</td>\n",
838
+ " <td>Jean-Pierre Guignon</td>\n",
839
+ " <td>2r3q57FhxdsCyYr0kuDq4b</td>\n",
840
+ " <td>track_26</td>\n",
841
+ " <td>artist</td>\n",
842
+ " </tr>\n",
843
+ " </tbody>\n",
844
+ "</table>\n",
845
+ "</div>\n",
846
+ " <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d959edf2-79c3-4736-8a38-326b6c0d0f2d')\"\n",
847
+ " title=\"Convert this dataframe to an interactive table.\"\n",
848
+ " style=\"display:none;\">\n",
849
+ " \n",
850
+ " <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
851
+ " width=\"24px\">\n",
852
+ " <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
853
+ " <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
854
+ " </svg>\n",
855
+ " </button>\n",
856
+ " \n",
857
+ " <style>\n",
858
+ " .colab-df-container {\n",
859
+ " display:flex;\n",
860
+ " flex-wrap:wrap;\n",
861
+ " gap: 12px;\n",
862
+ " }\n",
863
+ "\n",
864
+ " .colab-df-convert {\n",
865
+ " background-color: #E8F0FE;\n",
866
+ " border: none;\n",
867
+ " border-radius: 50%;\n",
868
+ " cursor: pointer;\n",
869
+ " display: none;\n",
870
+ " fill: #1967D2;\n",
871
+ " height: 32px;\n",
872
+ " padding: 0 0 0 0;\n",
873
+ " width: 32px;\n",
874
+ " }\n",
875
+ "\n",
876
+ " .colab-df-convert:hover {\n",
877
+ " background-color: #E2EBFA;\n",
878
+ " box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
879
+ " fill: #174EA6;\n",
880
+ " }\n",
881
+ "\n",
882
+ " [theme=dark] .colab-df-convert {\n",
883
+ " background-color: #3B4455;\n",
884
+ " fill: #D2E3FC;\n",
885
+ " }\n",
886
+ "\n",
887
+ " [theme=dark] .colab-df-convert:hover {\n",
888
+ " background-color: #434B5C;\n",
889
+ " box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
890
+ " filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
891
+ " fill: #FFFFFF;\n",
892
+ " }\n",
893
+ " </style>\n",
894
+ "\n",
895
+ " <script>\n",
896
+ " const buttonEl =\n",
897
+ " document.querySelector('#df-d959edf2-79c3-4736-8a38-326b6c0d0f2d button.colab-df-convert');\n",
898
+ " buttonEl.style.display =\n",
899
+ " google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
900
+ "\n",
901
+ " async function convertToInteractive(key) {\n",
902
+ " const element = document.querySelector('#df-d959edf2-79c3-4736-8a38-326b6c0d0f2d');\n",
903
+ " const dataTable =\n",
904
+ " await google.colab.kernel.invokeFunction('convertToInteractive',\n",
905
+ " [key], {});\n",
906
+ " if (!dataTable) return;\n",
907
+ "\n",
908
+ " const docLinkHtml = 'Like what you see? Visit the ' +\n",
909
+ " '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
910
+ " + ' to learn more about interactive tables.';\n",
911
+ " element.innerHTML = '';\n",
912
+ " dataTable['output_type'] = 'display_data';\n",
913
+ " await google.colab.output.renderOutput(dataTable, element);\n",
914
+ " const docLink = document.createElement('div');\n",
915
+ " docLink.innerHTML = docLinkHtml;\n",
916
+ " element.appendChild(docLink);\n",
917
+ " }\n",
918
+ " </script>\n",
919
+ " </div>\n",
920
+ " </div>\n",
921
+ " "
922
+ ]
923
+ },
924
+ "metadata": {}
925
+ },
926
+ {
927
+ "output_type": "execute_result",
928
+ "data": {
929
+ "text/plain": [
930
+ "Index(['Unnamed: 0', 'artist_popularity', 'followers', 'genres', 'id', 'name',\n",
931
+ " 'track_id', 'track_name_prev', 'type'],\n",
932
+ " dtype='object')"
933
+ ]
934
+ },
935
+ "metadata": {},
936
+ "execution_count": 13
937
+ }
938
+ ]
939
+ },
940
+ {
941
+ "cell_type": "code",
942
+ "source": [
943
+ "display(tracks_data.head())\n",
944
+ "tracks_data.columns"
945
+ ],
946
+ "metadata": {
947
+ "colab": {
948
+ "base_uri": "https://localhost:8080/",
949
+ "height": 594
950
+ },
951
+ "id": "OCjiZZnD6a0U",
952
+ "outputId": "3108dda2-b5a0-476d-9c1f-c61f58e8a396"
953
+ },
954
+ "execution_count": 14,
955
+ "outputs": [
956
+ {
957
+ "output_type": "display_data",
958
+ "data": {
959
+ "text/plain": [
960
+ " Unnamed: 0 acousticness album_id \\\n",
961
+ "0 0 0.294 0D3QufeCudpQANOR7luqdr \n",
962
+ "1 1 0.863 1bcqsH5UyTBzmh9YizdsBE \n",
963
+ "2 2 0.750 4tKijjmxGClg4JOLAyo2qE \n",
964
+ "3 3 0.763 6FeJF5r8roonnKraJxr4oB \n",
965
+ "4 4 0.770 4tKijjmxGClg4JOLAyo2qE \n",
966
+ "\n",
967
+ " analysis_url \\\n",
968
+ "0 https://api.spotify.com/v1/audio-analysis/5qlj... \n",
969
+ "1 https://api.spotify.com/v1/audio-analysis/3VAX... \n",
970
+ "2 https://api.spotify.com/v1/audio-analysis/1L3Y... \n",
971
+ "3 https://api.spotify.com/v1/audio-analysis/6aCe... \n",
972
+ "4 https://api.spotify.com/v1/audio-analysis/1Vo8... \n",
973
+ "\n",
974
+ " artists_id \\\n",
975
+ "0 ['3mxJuHRn2ZWD5OofvJtDZY'] \n",
976
+ "1 ['4xWMewm6CYMstu0sPgd9jJ'] \n",
977
+ "2 ['3hYaK5FF3YAglCj5HZgBnP'] \n",
978
+ "3 ['2KQsUB9DRBcJk17JWX1eXD'] \n",
979
+ "4 ['3hYaK5FF3YAglCj5HZgBnP'] \n",
980
+ "\n",
981
+ " available_markets country danceability \\\n",
982
+ "0 ['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH... BE 0.698 \n",
983
+ "1 ['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH... BE 0.719 \n",
984
+ "2 ['GB'] BE 0.466 \n",
985
+ "3 ['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH... BE 0.719 \n",
986
+ "4 ['GB'] BE 0.460 \n",
987
+ "\n",
988
+ " disc_number duration_ms ... \\\n",
989
+ "0 1.0 235584.0 ... \n",
990
+ "1 1.0 656960.0 ... \n",
991
+ "2 1.0 492840.0 ... \n",
992
+ "3 1.0 316578.0 ... \n",
993
+ "4 1.0 558880.0 ... \n",
994
+ "\n",
995
+ " preview_url speechiness tempo \\\n",
996
+ "0 https://p.scdn.co/mp3-preview/1b05a902da3a251d... 0.0262 115.018 \n",
997
+ "1 https://p.scdn.co/mp3-preview/d8140736a6131cb5... 0.9220 115.075 \n",
998
+ "2 https://p.scdn.co/mp3-preview/c8af28fb15185b18... 0.9440 79.565 \n",
999
+ "3 https://p.scdn.co/mp3-preview/7629b8e9f31f6e9b... 0.9380 112.822 \n",
1000
+ "4 https://p.scdn.co/mp3-preview/32be593c0eb82868... 0.9430 81.260 \n",
1001
+ "\n",
1002
+ " time_signature track_href \\\n",
1003
+ "0 4.0 https://api.spotify.com/v1/tracks/5qljLQuKnNJf... \n",
1004
+ "1 3.0 https://api.spotify.com/v1/tracks/3VAX2MJdmdqA... \n",
1005
+ "2 4.0 https://api.spotify.com/v1/tracks/1L3YAhsEMrGV... \n",
1006
+ "3 3.0 https://api.spotify.com/v1/tracks/6aCe9zzoZmCo... \n",
1007
+ "4 4.0 https://api.spotify.com/v1/tracks/1Vo802A38tPF... \n",
1008
+ "\n",
1009
+ " track_name_prev track_number uri \\\n",
1010
+ "0 track_14 1.0 spotify:track:5qljLQuKnNJf4F4vfxQB0V \n",
1011
+ "1 track_3 3.0 spotify:track:3VAX2MJdmdqARLSU5hPMpm \n",
1012
+ "2 track_4 4.0 spotify:track:1L3YAhsEMrGVvCgDXj2TYn \n",
1013
+ "3 track_9 1.0 spotify:track:6aCe9zzoZmCojX7bbgKKtf \n",
1014
+ "4 track_2 2.0 spotify:track:1Vo802A38tPFHmje1h91um \n",
1015
+ "\n",
1016
+ " valence type \n",
1017
+ "0 0.6220 track \n",
1018
+ "1 0.5890 track \n",
1019
+ "2 0.0850 track \n",
1020
+ "3 0.5330 track \n",
1021
+ "4 0.0906 track \n",
1022
+ "\n",
1023
+ "[5 rows x 32 columns]"
1024
+ ],
1025
+ "text/html": [
1026
+ "\n",
1027
+ " <div id=\"df-a317d767-9a8b-4a07-8bc0-8d0028475f92\">\n",
1028
+ " <div class=\"colab-df-container\">\n",
1029
+ " <div>\n",
1030
+ "<style scoped>\n",
1031
+ " .dataframe tbody tr th:only-of-type {\n",
1032
+ " vertical-align: middle;\n",
1033
+ " }\n",
1034
+ "\n",
1035
+ " .dataframe tbody tr th {\n",
1036
+ " vertical-align: top;\n",
1037
+ " }\n",
1038
+ "\n",
1039
+ " .dataframe thead th {\n",
1040
+ " text-align: right;\n",
1041
+ " }\n",
1042
+ "</style>\n",
1043
+ "<table border=\"1\" class=\"dataframe\">\n",
1044
+ " <thead>\n",
1045
+ " <tr style=\"text-align: right;\">\n",
1046
+ " <th></th>\n",
1047
+ " <th>Unnamed: 0</th>\n",
1048
+ " <th>acousticness</th>\n",
1049
+ " <th>album_id</th>\n",
1050
+ " <th>analysis_url</th>\n",
1051
+ " <th>artists_id</th>\n",
1052
+ " <th>available_markets</th>\n",
1053
+ " <th>country</th>\n",
1054
+ " <th>danceability</th>\n",
1055
+ " <th>disc_number</th>\n",
1056
+ " <th>duration_ms</th>\n",
1057
+ " <th>...</th>\n",
1058
+ " <th>preview_url</th>\n",
1059
+ " <th>speechiness</th>\n",
1060
+ " <th>tempo</th>\n",
1061
+ " <th>time_signature</th>\n",
1062
+ " <th>track_href</th>\n",
1063
+ " <th>track_name_prev</th>\n",
1064
+ " <th>track_number</th>\n",
1065
+ " <th>uri</th>\n",
1066
+ " <th>valence</th>\n",
1067
+ " <th>type</th>\n",
1068
+ " </tr>\n",
1069
+ " </thead>\n",
1070
+ " <tbody>\n",
1071
+ " <tr>\n",
1072
+ " <th>0</th>\n",
1073
+ " <td>0</td>\n",
1074
+ " <td>0.294</td>\n",
1075
+ " <td>0D3QufeCudpQANOR7luqdr</td>\n",
1076
+ " <td>https://api.spotify.com/v1/audio-analysis/5qlj...</td>\n",
1077
+ " <td>['3mxJuHRn2ZWD5OofvJtDZY']</td>\n",
1078
+ " <td>['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH...</td>\n",
1079
+ " <td>BE</td>\n",
1080
+ " <td>0.698</td>\n",
1081
+ " <td>1.0</td>\n",
1082
+ " <td>235584.0</td>\n",
1083
+ " <td>...</td>\n",
1084
+ " <td>https://p.scdn.co/mp3-preview/1b05a902da3a251d...</td>\n",
1085
+ " <td>0.0262</td>\n",
1086
+ " <td>115.018</td>\n",
1087
+ " <td>4.0</td>\n",
1088
+ " <td>https://api.spotify.com/v1/tracks/5qljLQuKnNJf...</td>\n",
1089
+ " <td>track_14</td>\n",
1090
+ " <td>1.0</td>\n",
1091
+ " <td>spotify:track:5qljLQuKnNJf4F4vfxQB0V</td>\n",
1092
+ " <td>0.6220</td>\n",
1093
+ " <td>track</td>\n",
1094
+ " </tr>\n",
1095
+ " <tr>\n",
1096
+ " <th>1</th>\n",
1097
+ " <td>1</td>\n",
1098
+ " <td>0.863</td>\n",
1099
+ " <td>1bcqsH5UyTBzmh9YizdsBE</td>\n",
1100
+ " <td>https://api.spotify.com/v1/audio-analysis/3VAX...</td>\n",
1101
+ " <td>['4xWMewm6CYMstu0sPgd9jJ']</td>\n",
1102
+ " <td>['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH...</td>\n",
1103
+ " <td>BE</td>\n",
1104
+ " <td>0.719</td>\n",
1105
+ " <td>1.0</td>\n",
1106
+ " <td>656960.0</td>\n",
1107
+ " <td>...</td>\n",
1108
+ " <td>https://p.scdn.co/mp3-preview/d8140736a6131cb5...</td>\n",
1109
+ " <td>0.9220</td>\n",
1110
+ " <td>115.075</td>\n",
1111
+ " <td>3.0</td>\n",
1112
+ " <td>https://api.spotify.com/v1/tracks/3VAX2MJdmdqA...</td>\n",
1113
+ " <td>track_3</td>\n",
1114
+ " <td>3.0</td>\n",
1115
+ " <td>spotify:track:3VAX2MJdmdqARLSU5hPMpm</td>\n",
1116
+ " <td>0.5890</td>\n",
1117
+ " <td>track</td>\n",
1118
+ " </tr>\n",
1119
+ " <tr>\n",
1120
+ " <th>2</th>\n",
1121
+ " <td>2</td>\n",
1122
+ " <td>0.750</td>\n",
1123
+ " <td>4tKijjmxGClg4JOLAyo2qE</td>\n",
1124
+ " <td>https://api.spotify.com/v1/audio-analysis/1L3Y...</td>\n",
1125
+ " <td>['3hYaK5FF3YAglCj5HZgBnP']</td>\n",
1126
+ " <td>['GB']</td>\n",
1127
+ " <td>BE</td>\n",
1128
+ " <td>0.466</td>\n",
1129
+ " <td>1.0</td>\n",
1130
+ " <td>492840.0</td>\n",
1131
+ " <td>...</td>\n",
1132
+ " <td>https://p.scdn.co/mp3-preview/c8af28fb15185b18...</td>\n",
1133
+ " <td>0.9440</td>\n",
1134
+ " <td>79.565</td>\n",
1135
+ " <td>4.0</td>\n",
1136
+ " <td>https://api.spotify.com/v1/tracks/1L3YAhsEMrGV...</td>\n",
1137
+ " <td>track_4</td>\n",
1138
+ " <td>4.0</td>\n",
1139
+ " <td>spotify:track:1L3YAhsEMrGVvCgDXj2TYn</td>\n",
1140
+ " <td>0.0850</td>\n",
1141
+ " <td>track</td>\n",
1142
+ " </tr>\n",
1143
+ " <tr>\n",
1144
+ " <th>3</th>\n",
1145
+ " <td>3</td>\n",
1146
+ " <td>0.763</td>\n",
1147
+ " <td>6FeJF5r8roonnKraJxr4oB</td>\n",
1148
+ " <td>https://api.spotify.com/v1/audio-analysis/6aCe...</td>\n",
1149
+ " <td>['2KQsUB9DRBcJk17JWX1eXD']</td>\n",
1150
+ " <td>['AD', 'AE', 'AR', 'AT', 'AU', 'BE', 'BG', 'BH...</td>\n",
1151
+ " <td>BE</td>\n",
1152
+ " <td>0.719</td>\n",
1153
+ " <td>1.0</td>\n",
1154
+ " <td>316578.0</td>\n",
1155
+ " <td>...</td>\n",
1156
+ " <td>https://p.scdn.co/mp3-preview/7629b8e9f31f6e9b...</td>\n",
1157
+ " <td>0.9380</td>\n",
1158
+ " <td>112.822</td>\n",
1159
+ " <td>3.0</td>\n",
1160
+ " <td>https://api.spotify.com/v1/tracks/6aCe9zzoZmCo...</td>\n",
1161
+ " <td>track_9</td>\n",
1162
+ " <td>1.0</td>\n",
1163
+ " <td>spotify:track:6aCe9zzoZmCojX7bbgKKtf</td>\n",
1164
+ " <td>0.5330</td>\n",
1165
+ " <td>track</td>\n",
1166
+ " </tr>\n",
1167
+ " <tr>\n",
1168
+ " <th>4</th>\n",
1169
+ " <td>4</td>\n",
1170
+ " <td>0.770</td>\n",
1171
+ " <td>4tKijjmxGClg4JOLAyo2qE</td>\n",
1172
+ " <td>https://api.spotify.com/v1/audio-analysis/1Vo8...</td>\n",
1173
+ " <td>['3hYaK5FF3YAglCj5HZgBnP']</td>\n",
1174
+ " <td>['GB']</td>\n",
1175
+ " <td>BE</td>\n",
1176
+ " <td>0.460</td>\n",
1177
+ " <td>1.0</td>\n",
1178
+ " <td>558880.0</td>\n",
1179
+ " <td>...</td>\n",
1180
+ " <td>https://p.scdn.co/mp3-preview/32be593c0eb82868...</td>\n",
1181
+ " <td>0.9430</td>\n",
1182
+ " <td>81.260</td>\n",
1183
+ " <td>4.0</td>\n",
1184
+ " <td>https://api.spotify.com/v1/tracks/1Vo802A38tPF...</td>\n",
1185
+ " <td>track_2</td>\n",
1186
+ " <td>2.0</td>\n",
1187
+ " <td>spotify:track:1Vo802A38tPFHmje1h91um</td>\n",
1188
+ " <td>0.0906</td>\n",
1189
+ " <td>track</td>\n",
1190
+ " </tr>\n",
1191
+ " </tbody>\n",
1192
+ "</table>\n",
1193
+ "<p>5 rows × 32 columns</p>\n",
1194
+ "</div>\n",
1195
+ " <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a317d767-9a8b-4a07-8bc0-8d0028475f92')\"\n",
1196
+ " title=\"Convert this dataframe to an interactive table.\"\n",
1197
+ " style=\"display:none;\">\n",
1198
+ " \n",
1199
+ " <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
1200
+ " width=\"24px\">\n",
1201
+ " <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
1202
+ " <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
1203
+ " </svg>\n",
1204
+ " </button>\n",
1205
+ " \n",
1206
+ " <style>\n",
1207
+ " .colab-df-container {\n",
1208
+ " display:flex;\n",
1209
+ " flex-wrap:wrap;\n",
1210
+ " gap: 12px;\n",
1211
+ " }\n",
1212
+ "\n",
1213
+ " .colab-df-convert {\n",
1214
+ " background-color: #E8F0FE;\n",
1215
+ " border: none;\n",
1216
+ " border-radius: 50%;\n",
1217
+ " cursor: pointer;\n",
1218
+ " display: none;\n",
1219
+ " fill: #1967D2;\n",
1220
+ " height: 32px;\n",
1221
+ " padding: 0 0 0 0;\n",
1222
+ " width: 32px;\n",
1223
+ " }\n",
1224
+ "\n",
1225
+ " .colab-df-convert:hover {\n",
1226
+ " background-color: #E2EBFA;\n",
1227
+ " box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
1228
+ " fill: #174EA6;\n",
1229
+ " }\n",
1230
+ "\n",
1231
+ " [theme=dark] .colab-df-convert {\n",
1232
+ " background-color: #3B4455;\n",
1233
+ " fill: #D2E3FC;\n",
1234
+ " }\n",
1235
+ "\n",
1236
+ " [theme=dark] .colab-df-convert:hover {\n",
1237
+ " background-color: #434B5C;\n",
1238
+ " box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
1239
+ " filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
1240
+ " fill: #FFFFFF;\n",
1241
+ " }\n",
1242
+ " </style>\n",
1243
+ "\n",
1244
+ " <script>\n",
1245
+ " const buttonEl =\n",
1246
+ " document.querySelector('#df-a317d767-9a8b-4a07-8bc0-8d0028475f92 button.colab-df-convert');\n",
1247
+ " buttonEl.style.display =\n",
1248
+ " google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
1249
+ "\n",
1250
+ " async function convertToInteractive(key) {\n",
1251
+ " const element = document.querySelector('#df-a317d767-9a8b-4a07-8bc0-8d0028475f92');\n",
1252
+ " const dataTable =\n",
1253
+ " await google.colab.kernel.invokeFunction('convertToInteractive',\n",
1254
+ " [key], {});\n",
1255
+ " if (!dataTable) return;\n",
1256
+ "\n",
1257
+ " const docLinkHtml = 'Like what you see? Visit the ' +\n",
1258
+ " '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
1259
+ " + ' to learn more about interactive tables.';\n",
1260
+ " element.innerHTML = '';\n",
1261
+ " dataTable['output_type'] = 'display_data';\n",
1262
+ " await google.colab.output.renderOutput(dataTable, element);\n",
1263
+ " const docLink = document.createElement('div');\n",
1264
+ " docLink.innerHTML = docLinkHtml;\n",
1265
+ " element.appendChild(docLink);\n",
1266
+ " }\n",
1267
+ " </script>\n",
1268
+ " </div>\n",
1269
+ " </div>\n",
1270
+ " "
1271
+ ]
1272
+ },
1273
+ "metadata": {}
1274
+ },
1275
+ {
1276
+ "output_type": "execute_result",
1277
+ "data": {
1278
+ "text/plain": [
1279
+ "Index(['Unnamed: 0', 'acousticness', 'album_id', 'analysis_url', 'artists_id',\n",
1280
+ " 'available_markets', 'country', 'danceability', 'disc_number',\n",
1281
+ " 'duration_ms', 'energy', 'href', 'id', 'instrumentalness', 'key',\n",
1282
+ " 'liveness', 'loudness', 'lyrics', 'mode', 'name', 'playlist',\n",
1283
+ " 'popularity', 'preview_url', 'speechiness', 'tempo', 'time_signature',\n",
1284
+ " 'track_href', 'track_name_prev', 'track_number', 'uri', 'valence',\n",
1285
+ " 'type'],\n",
1286
+ " dtype='object')"
1287
+ ]
1288
+ },
1289
+ "metadata": {},
1290
+ "execution_count": 14
1291
+ }
1292
+ ]
1293
+ },
1294
+ {
1295
+ "cell_type": "code",
1296
+ "source": [
1297
+ "## join artist genre information and album release date with track dataset\n",
1298
+ "# drop irrelevant columns\n",
1299
+ "# get only tracks after 1990\n",
1300
+ "def join_genre_and_date(artist_df, album_df, track_df):\n",
1301
+ " album = album_df.rename(columns={'id':\"album_id\"}).set_index('album_id')\n",
1302
+ " artist = artist_df.rename(columns={'id':\"artists_id\",'name':\"artists_name\"}).set_index('artists_id')\n",
1303
+ " track = track_df.set_index('album_id').join(album['release_date'], on='album_id' )\n",
1304
+ " track.artists_id = track.artists_id.apply(lambda x: x[2:-2])\n",
1305
+ " track = track.set_index('artists_id').join(artist[['artists_name','genres']], on='artists_id' )\n",
1306
+ " track.reset_index(drop=False, inplace=True)\n",
1307
+ " track['release_year'] = pd.to_datetime(track.release_date).dt.year\n",
1308
+ " track.drop(columns = ['Unnamed: 0','country','track_name_prev','track_number','type'], inplace = True)\n",
1309
+ " \n",
1310
+ " return track[track.release_year >= 1990]"
1311
+ ],
1312
+ "metadata": {
1313
+ "id": "LZZhGSn46c_y"
1314
+ },
1315
+ "execution_count": 15,
1316
+ "outputs": []
1317
+ },
1318
+ {
1319
+ "cell_type": "code",
1320
+ "source": [
1321
+ "def get_filtered_track_df(df, genres_to_include):\n",
1322
+ " df['genres'] = df.genres.apply(lambda x: [i[1:-1] for i in str(x)[1:-1].split(\", \")])\n",
1323
+ " df_exploded = df.explode(\"genres\")[df.explode(\"genres\")[\"genres\"].isin(genres_to_include)]\n",
1324
+ " df_exploded.loc[df_exploded[\"genres\"]==\"korean pop\", \"genres\"] = \"k-pop\"\n",
1325
+ " df_exploded_indices = list(df_exploded.index.unique())\n",
1326
+ " df = df[df.index.isin(df_exploded_indices)]\n",
1327
+ " df = df.reset_index(drop=True)\n",
1328
+ " return df"
1329
+ ],
1330
+ "metadata": {
1331
+ "id": "rP905-I66eB0"
1332
+ },
1333
+ "execution_count": 16,
1334
+ "outputs": []
1335
+ },
1336
+ {
1337
+ "cell_type": "code",
1338
+ "source": [
1339
+ "track_with_year_and_genre = join_genre_and_date(artists_data, albums_data, tracks_data)\n",
1340
+ "genres_to_include = genres = ['dance pop', 'electronic', 'electropop', 'hip hop', 'jazz', 'k-pop', 'latin', 'pop', 'pop rap', 'r&b', 'rock']\n",
1341
+ "filtered_track_df = get_filtered_track_df(track_with_year_and_genre, genres_to_include)"
1342
+ ],
1343
+ "metadata": {
1344
+ "id": "7BmEttOn6iWy"
1345
+ },
1346
+ "execution_count": 18,
1347
+ "outputs": []
1348
+ },
1349
+ {
1350
+ "cell_type": "code",
1351
+ "source": [
1352
+ "filtered_track_df[\"uri\"] = filtered_track_df[\"uri\"].str.replace(\"spotify:track:\", \"\")\n",
1353
+ "filtered_track_df = filtered_track_df.drop(columns=['analysis_url', 'available_markets'])"
1354
+ ],
1355
+ "metadata": {
1356
+ "id": "ghqON6di6n3e"
1357
+ },
1358
+ "execution_count": 19,
1359
+ "outputs": []
1360
+ },
1361
+ {
1362
+ "cell_type": "code",
1363
+ "source": [
1364
+ "display(filtered_track_df.head())\n",
1365
+ "filtered_track_df.columns"
1366
+ ],
1367
+ "metadata": {
1368
+ "colab": {
1369
+ "base_uri": "https://localhost:8080/",
1370
+ "height": 751
1371
+ },
1372
+ "id": "a3Hwe1mB6qMU",
1373
+ "outputId": "0ed05171-2d9d-4b56-f3e3-b04e14485c79"
1374
+ },
1375
+ "execution_count": 20,
1376
+ "outputs": [
1377
+ {
1378
+ "output_type": "display_data",
1379
+ "data": {
1380
+ "text/plain": [
1381
+ " artists_id acousticness danceability disc_number \\\n",
1382
+ "0 68WwJXWrpo1yVOOIZjLSeT 0.0268 0.506 1.0 \n",
1383
+ "1 09xj0S68Y1OU1vHMCZAIvz 0.5050 0.487 1.0 \n",
1384
+ "2 6pSsE5y0uJMwYj83KrPyf9 0.1330 0.629 1.0 \n",
1385
+ "3 7slfeZO9LsJbWgpkIoXBUJ 0.4060 0.590 1.0 \n",
1386
+ "4 09hVIj6vWgoCDtT03h8ZCa 0.0316 0.727 1.0 \n",
1387
+ "\n",
1388
+ " duration_ms energy href \\\n",
1389
+ "0 248777.0 0.741 https://api.spotify.com/v1/tracks/0UATU9OJxh4m... \n",
1390
+ "1 171573.0 0.297 https://api.spotify.com/v1/tracks/4JH1M62gVDND... \n",
1391
+ "2 207396.0 0.706 https://api.spotify.com/v1/tracks/0h7Ld5CvgzaU... \n",
1392
+ "3 279000.0 0.597 https://api.spotify.com/v1/tracks/4S1bYWrLOC8s... \n",
1393
+ "4 218773.0 0.380 https://api.spotify.com/v1/tracks/758mQT4zzlvB... \n",
1394
+ "\n",
1395
+ " id instrumentalness key ... speechiness tempo \\\n",
1396
+ "0 0UATU9OJxh4m3fwDljdGZn 0.000027 1.0 ... 0.0349 94.042 \n",
1397
+ "1 4JH1M62gVDNDhDAUiQB3Qv 0.000052 11.0 ... 0.0915 185.912 \n",
1398
+ "2 0h7Ld5CvgzaUN1zA3tdyPq 0.000000 1.0 ... 0.4360 81.220 \n",
1399
+ "3 4S1bYWrLOC8smuy8kJzxKQ 0.000023 9.0 ... 0.0275 121.051 \n",
1400
+ "4 758mQT4zzlvBhy9PvNePwC 0.000000 7.0 ... 0.3350 92.050 \n",
1401
+ "\n",
1402
+ " time_signature track_href \\\n",
1403
+ "0 4.0 https://api.spotify.com/v1/tracks/0UATU9OJxh4m... \n",
1404
+ "1 3.0 https://api.spotify.com/v1/tracks/4JH1M62gVDND... \n",
1405
+ "2 4.0 https://api.spotify.com/v1/tracks/0h7Ld5CvgzaU... \n",
1406
+ "3 4.0 https://api.spotify.com/v1/tracks/4S1bYWrLOC8s... \n",
1407
+ "4 4.0 https://api.spotify.com/v1/tracks/758mQT4zzlvB... \n",
1408
+ "\n",
1409
+ " uri valence release_date artists_name \\\n",
1410
+ "0 0UATU9OJxh4m3fwDljdGZn 0.236 2018-09-28 Evalyn \n",
1411
+ "1 4JH1M62gVDNDhDAUiQB3Qv 0.289 2001-08-21 Café Tacvba \n",
1412
+ "2 0h7Ld5CvgzaUN1zA3tdyPq 0.543 2019-01-25 Dawn Richard \n",
1413
+ "3 4S1bYWrLOC8smuy8kJzxKQ 0.466 1995-09-12 Ricky Martin \n",
1414
+ "4 758mQT4zzlvBhy9PvNePwC 0.455 1991-09-24 A Tribe Called Quest \n",
1415
+ "\n",
1416
+ " genres release_year \n",
1417
+ "0 [electropop, indie electro-pop, indie poptimis... 2018 \n",
1418
+ "1 [latin, latin alternative, latin rock, mexican... 2001 \n",
1419
+ "2 [alternative r&b, deep pop r&b, escape room, h... 2019 \n",
1420
+ "3 [dance pop, latin, latin pop, mexican pop, pop... 1995 \n",
1421
+ "4 [alternative hip hop, conscious hip hop, east ... 1991 \n",
1422
+ "\n",
1423
+ "[5 rows x 28 columns]"
1424
+ ],
1425
+ "text/html": [
1426
+ "\n",
1427
+ " <div id=\"df-2ac09e30-b225-407d-9e95-1016823bdae8\">\n",
1428
+ " <div class=\"colab-df-container\">\n",
1429
+ " <div>\n",
1430
+ "<style scoped>\n",
1431
+ " .dataframe tbody tr th:only-of-type {\n",
1432
+ " vertical-align: middle;\n",
1433
+ " }\n",
1434
+ "\n",
1435
+ " .dataframe tbody tr th {\n",
1436
+ " vertical-align: top;\n",
1437
+ " }\n",
1438
+ "\n",
1439
+ " .dataframe thead th {\n",
1440
+ " text-align: right;\n",
1441
+ " }\n",
1442
+ "</style>\n",
1443
+ "<table border=\"1\" class=\"dataframe\">\n",
1444
+ " <thead>\n",
1445
+ " <tr style=\"text-align: right;\">\n",
1446
+ " <th></th>\n",
1447
+ " <th>artists_id</th>\n",
1448
+ " <th>acousticness</th>\n",
1449
+ " <th>danceability</th>\n",
1450
+ " <th>disc_number</th>\n",
1451
+ " <th>duration_ms</th>\n",
1452
+ " <th>energy</th>\n",
1453
+ " <th>href</th>\n",
1454
+ " <th>id</th>\n",
1455
+ " <th>instrumentalness</th>\n",
1456
+ " <th>key</th>\n",
1457
+ " <th>...</th>\n",
1458
+ " <th>speechiness</th>\n",
1459
+ " <th>tempo</th>\n",
1460
+ " <th>time_signature</th>\n",
1461
+ " <th>track_href</th>\n",
1462
+ " <th>uri</th>\n",
1463
+ " <th>valence</th>\n",
1464
+ " <th>release_date</th>\n",
1465
+ " <th>artists_name</th>\n",
1466
+ " <th>genres</th>\n",
1467
+ " <th>release_year</th>\n",
1468
+ " </tr>\n",
1469
+ " </thead>\n",
1470
+ " <tbody>\n",
1471
+ " <tr>\n",
1472
+ " <th>0</th>\n",
1473
+ " <td>68WwJXWrpo1yVOOIZjLSeT</td>\n",
1474
+ " <td>0.0268</td>\n",
1475
+ " <td>0.506</td>\n",
1476
+ " <td>1.0</td>\n",
1477
+ " <td>248777.0</td>\n",
1478
+ " <td>0.741</td>\n",
1479
+ " <td>https://api.spotify.com/v1/tracks/0UATU9OJxh4m...</td>\n",
1480
+ " <td>0UATU9OJxh4m3fwDljdGZn</td>\n",
1481
+ " <td>0.000027</td>\n",
1482
+ " <td>1.0</td>\n",
1483
+ " <td>...</td>\n",
1484
+ " <td>0.0349</td>\n",
1485
+ " <td>94.042</td>\n",
1486
+ " <td>4.0</td>\n",
1487
+ " <td>https://api.spotify.com/v1/tracks/0UATU9OJxh4m...</td>\n",
1488
+ " <td>0UATU9OJxh4m3fwDljdGZn</td>\n",
1489
+ " <td>0.236</td>\n",
1490
+ " <td>2018-09-28</td>\n",
1491
+ " <td>Evalyn</td>\n",
1492
+ " <td>[electropop, indie electro-pop, indie poptimis...</td>\n",
1493
+ " <td>2018</td>\n",
1494
+ " </tr>\n",
1495
+ " <tr>\n",
1496
+ " <th>1</th>\n",
1497
+ " <td>09xj0S68Y1OU1vHMCZAIvz</td>\n",
1498
+ " <td>0.5050</td>\n",
1499
+ " <td>0.487</td>\n",
1500
+ " <td>1.0</td>\n",
1501
+ " <td>171573.0</td>\n",
1502
+ " <td>0.297</td>\n",
1503
+ " <td>https://api.spotify.com/v1/tracks/4JH1M62gVDND...</td>\n",
1504
+ " <td>4JH1M62gVDNDhDAUiQB3Qv</td>\n",
1505
+ " <td>0.000052</td>\n",
1506
+ " <td>11.0</td>\n",
1507
+ " <td>...</td>\n",
1508
+ " <td>0.0915</td>\n",
1509
+ " <td>185.912</td>\n",
1510
+ " <td>3.0</td>\n",
1511
+ " <td>https://api.spotify.com/v1/tracks/4JH1M62gVDND...</td>\n",
1512
+ " <td>4JH1M62gVDNDhDAUiQB3Qv</td>\n",
1513
+ " <td>0.289</td>\n",
1514
+ " <td>2001-08-21</td>\n",
1515
+ " <td>Café Tacvba</td>\n",
1516
+ " <td>[latin, latin alternative, latin rock, mexican...</td>\n",
1517
+ " <td>2001</td>\n",
1518
+ " </tr>\n",
1519
+ " <tr>\n",
1520
+ " <th>2</th>\n",
1521
+ " <td>6pSsE5y0uJMwYj83KrPyf9</td>\n",
1522
+ " <td>0.1330</td>\n",
1523
+ " <td>0.629</td>\n",
1524
+ " <td>1.0</td>\n",
1525
+ " <td>207396.0</td>\n",
1526
+ " <td>0.706</td>\n",
1527
+ " <td>https://api.spotify.com/v1/tracks/0h7Ld5CvgzaU...</td>\n",
1528
+ " <td>0h7Ld5CvgzaUN1zA3tdyPq</td>\n",
1529
+ " <td>0.000000</td>\n",
1530
+ " <td>1.0</td>\n",
1531
+ " <td>...</td>\n",
1532
+ " <td>0.4360</td>\n",
1533
+ " <td>81.220</td>\n",
1534
+ " <td>4.0</td>\n",
1535
+ " <td>https://api.spotify.com/v1/tracks/0h7Ld5CvgzaU...</td>\n",
1536
+ " <td>0h7Ld5CvgzaUN1zA3tdyPq</td>\n",
1537
+ " <td>0.543</td>\n",
1538
+ " <td>2019-01-25</td>\n",
1539
+ " <td>Dawn Richard</td>\n",
1540
+ " <td>[alternative r&amp;b, deep pop r&amp;b, escape room, h...</td>\n",
1541
+ " <td>2019</td>\n",
1542
+ " </tr>\n",
1543
+ " <tr>\n",
1544
+ " <th>3</th>\n",
1545
+ " <td>7slfeZO9LsJbWgpkIoXBUJ</td>\n",
1546
+ " <td>0.4060</td>\n",
1547
+ " <td>0.590</td>\n",
1548
+ " <td>1.0</td>\n",
1549
+ " <td>279000.0</td>\n",
1550
+ " <td>0.597</td>\n",
1551
+ " <td>https://api.spotify.com/v1/tracks/4S1bYWrLOC8s...</td>\n",
1552
+ " <td>4S1bYWrLOC8smuy8kJzxKQ</td>\n",
1553
+ " <td>0.000023</td>\n",
1554
+ " <td>9.0</td>\n",
1555
+ " <td>...</td>\n",
1556
+ " <td>0.0275</td>\n",
1557
+ " <td>121.051</td>\n",
1558
+ " <td>4.0</td>\n",
1559
+ " <td>https://api.spotify.com/v1/tracks/4S1bYWrLOC8s...</td>\n",
1560
+ " <td>4S1bYWrLOC8smuy8kJzxKQ</td>\n",
1561
+ " <td>0.466</td>\n",
1562
+ " <td>1995-09-12</td>\n",
1563
+ " <td>Ricky Martin</td>\n",
1564
+ " <td>[dance pop, latin, latin pop, mexican pop, pop...</td>\n",
1565
+ " <td>1995</td>\n",
1566
+ " </tr>\n",
1567
+ " <tr>\n",
1568
+ " <th>4</th>\n",
1569
+ " <td>09hVIj6vWgoCDtT03h8ZCa</td>\n",
1570
+ " <td>0.0316</td>\n",
1571
+ " <td>0.727</td>\n",
1572
+ " <td>1.0</td>\n",
1573
+ " <td>218773.0</td>\n",
1574
+ " <td>0.380</td>\n",
1575
+ " <td>https://api.spotify.com/v1/tracks/758mQT4zzlvB...</td>\n",
1576
+ " <td>758mQT4zzlvBhy9PvNePwC</td>\n",
1577
+ " <td>0.000000</td>\n",
1578
+ " <td>7.0</td>\n",
1579
+ " <td>...</td>\n",
1580
+ " <td>0.3350</td>\n",
1581
+ " <td>92.050</td>\n",
1582
+ " <td>4.0</td>\n",
1583
+ " <td>https://api.spotify.com/v1/tracks/758mQT4zzlvB...</td>\n",
1584
+ " <td>758mQT4zzlvBhy9PvNePwC</td>\n",
1585
+ " <td>0.455</td>\n",
1586
+ " <td>1991-09-24</td>\n",
1587
+ " <td>A Tribe Called Quest</td>\n",
1588
+ " <td>[alternative hip hop, conscious hip hop, east ...</td>\n",
1589
+ " <td>1991</td>\n",
1590
+ " </tr>\n",
1591
+ " </tbody>\n",
1592
+ "</table>\n",
1593
+ "<p>5 rows × 28 columns</p>\n",
1594
+ "</div>\n",
1595
+ " <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-2ac09e30-b225-407d-9e95-1016823bdae8')\"\n",
1596
+ " title=\"Convert this dataframe to an interactive table.\"\n",
1597
+ " style=\"display:none;\">\n",
1598
+ " \n",
1599
+ " <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
1600
+ " width=\"24px\">\n",
1601
+ " <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
1602
+ " <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
1603
+ " </svg>\n",
1604
+ " </button>\n",
1605
+ " \n",
1606
+ " <style>\n",
1607
+ " .colab-df-container {\n",
1608
+ " display:flex;\n",
1609
+ " flex-wrap:wrap;\n",
1610
+ " gap: 12px;\n",
1611
+ " }\n",
1612
+ "\n",
1613
+ " .colab-df-convert {\n",
1614
+ " background-color: #E8F0FE;\n",
1615
+ " border: none;\n",
1616
+ " border-radius: 50%;\n",
1617
+ " cursor: pointer;\n",
1618
+ " display: none;\n",
1619
+ " fill: #1967D2;\n",
1620
+ " height: 32px;\n",
1621
+ " padding: 0 0 0 0;\n",
1622
+ " width: 32px;\n",
1623
+ " }\n",
1624
+ "\n",
1625
+ " .colab-df-convert:hover {\n",
1626
+ " background-color: #E2EBFA;\n",
1627
+ " box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
1628
+ " fill: #174EA6;\n",
1629
+ " }\n",
1630
+ "\n",
1631
+ " [theme=dark] .colab-df-convert {\n",
1632
+ " background-color: #3B4455;\n",
1633
+ " fill: #D2E3FC;\n",
1634
+ " }\n",
1635
+ "\n",
1636
+ " [theme=dark] .colab-df-convert:hover {\n",
1637
+ " background-color: #434B5C;\n",
1638
+ " box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
1639
+ " filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
1640
+ " fill: #FFFFFF;\n",
1641
+ " }\n",
1642
+ " </style>\n",
1643
+ "\n",
1644
+ " <script>\n",
1645
+ " const buttonEl =\n",
1646
+ " document.querySelector('#df-2ac09e30-b225-407d-9e95-1016823bdae8 button.colab-df-convert');\n",
1647
+ " buttonEl.style.display =\n",
1648
+ " google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
1649
+ "\n",
1650
+ " async function convertToInteractive(key) {\n",
1651
+ " const element = document.querySelector('#df-2ac09e30-b225-407d-9e95-1016823bdae8');\n",
1652
+ " const dataTable =\n",
1653
+ " await google.colab.kernel.invokeFunction('convertToInteractive',\n",
1654
+ " [key], {});\n",
1655
+ " if (!dataTable) return;\n",
1656
+ "\n",
1657
+ " const docLinkHtml = 'Like what you see? Visit the ' +\n",
1658
+ " '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
1659
+ " + ' to learn more about interactive tables.';\n",
1660
+ " element.innerHTML = '';\n",
1661
+ " dataTable['output_type'] = 'display_data';\n",
1662
+ " await google.colab.output.renderOutput(dataTable, element);\n",
1663
+ " const docLink = document.createElement('div');\n",
1664
+ " docLink.innerHTML = docLinkHtml;\n",
1665
+ " element.appendChild(docLink);\n",
1666
+ " }\n",
1667
+ " </script>\n",
1668
+ " </div>\n",
1669
+ " </div>\n",
1670
+ " "
1671
+ ]
1672
+ },
1673
+ "metadata": {}
1674
+ },
1675
+ {
1676
+ "output_type": "execute_result",
1677
+ "data": {
1678
+ "text/plain": [
1679
+ "Index(['artists_id', 'acousticness', 'danceability', 'disc_number',\n",
1680
+ " 'duration_ms', 'energy', 'href', 'id', 'instrumentalness', 'key',\n",
1681
+ " 'liveness', 'loudness', 'lyrics', 'mode', 'name', 'playlist',\n",
1682
+ " 'popularity', 'preview_url', 'speechiness', 'tempo', 'time_signature',\n",
1683
+ " 'track_href', 'uri', 'valence', 'release_date', 'artists_name',\n",
1684
+ " 'genres', 'release_year'],\n",
1685
+ " dtype='object')"
1686
+ ]
1687
+ },
1688
+ "metadata": {},
1689
+ "execution_count": 20
1690
+ }
1691
+ ]
1692
+ },
1693
+ {
1694
+ "cell_type": "code",
1695
+ "source": [
1696
+ "filtered_track_df.to_csv(\"filtered_track_df.csv\", index=False)"
1697
+ ],
1698
+ "metadata": {
1699
+ "id": "O9jsAS2D6rqY"
1700
+ },
1701
+ "execution_count": 21,
1702
+ "outputs": []
1703
+ },
1704
+ {
1705
+ "cell_type": "code",
1706
+ "source": [
1707
+ "tracks_data.describe()"
1708
+ ],
1709
+ "metadata": {
1710
+ "colab": {
1711
+ "base_uri": "https://localhost:8080/",
1712
+ "height": 364
1713
+ },
1714
+ "id": "T7MIXjFh6urm",
1715
+ "outputId": "9ae43e81-94a4-4cfc-8be2-3c2e30ea0c62"
1716
+ },
1717
+ "execution_count": 22,
1718
+ "outputs": [
1719
+ {
1720
+ "output_type": "execute_result",
1721
+ "data": {
1722
+ "text/plain": [
1723
+ " Unnamed: 0 acousticness danceability disc_number \\\n",
1724
+ "count 101939.000000 101939.000000 101939.000000 101939.000000 \n",
1725
+ "mean 50969.000000 0.352124 0.586015 1.032166 \n",
1726
+ "std 29427.398883 0.334855 0.177724 0.566789 \n",
1727
+ "min 0.000000 0.000000 0.000000 1.000000 \n",
1728
+ "25% 25484.500000 0.040700 0.480000 1.000000 \n",
1729
+ "50% 50969.000000 0.238000 0.610000 1.000000 \n",
1730
+ "75% 76453.500000 0.645000 0.714000 1.000000 \n",
1731
+ "max 101938.000000 0.996000 0.989000 81.000000 \n",
1732
+ "\n",
1733
+ " duration_ms energy instrumentalness key \\\n",
1734
+ "count 1.019390e+05 101939.000000 101939.000000 101939.000000 \n",
1735
+ "mean 2.467708e+05 0.586479 0.148776 5.270858 \n",
1736
+ "std 1.904303e+05 0.260170 0.304024 3.577679 \n",
1737
+ "min 1.155000e+03 0.000000 0.000000 0.000000 \n",
1738
+ "25% 1.840000e+05 0.411000 0.000000 2.000000 \n",
1739
+ "50% 2.168930e+05 0.629000 0.000037 5.000000 \n",
1740
+ "75% 2.610550e+05 0.798000 0.034400 8.000000 \n",
1741
+ "max 5.505831e+06 1.000000 1.000000 11.000000 \n",
1742
+ "\n",
1743
+ " liveness loudness mode popularity \\\n",
1744
+ "count 101939.000000 101939.000000 101939.000000 101939.000000 \n",
1745
+ "mean 0.197640 -9.462720 0.618154 39.782311 \n",
1746
+ "std 0.175391 6.198508 0.485841 16.790769 \n",
1747
+ "min 0.000000 -60.000000 0.000000 0.000000 \n",
1748
+ "25% 0.095600 -11.149000 0.000000 29.000000 \n",
1749
+ "50% 0.124000 -7.599000 1.000000 41.000000 \n",
1750
+ "75% 0.241000 -5.509000 1.000000 52.000000 \n",
1751
+ "max 0.999000 2.719000 1.000000 97.000000 \n",
1752
+ "\n",
1753
+ " speechiness tempo time_signature track_number \\\n",
1754
+ "count 101939.000000 101939.000000 101939.000000 101939.000000 \n",
1755
+ "mean 0.128841 118.358527 3.875651 4.608060 \n",
1756
+ "std 0.203324 30.224074 0.517008 7.181805 \n",
1757
+ "min 0.000000 0.000000 0.000000 1.000000 \n",
1758
+ "25% 0.036400 95.973000 4.000000 1.000000 \n",
1759
+ "50% 0.050600 118.067000 4.000000 2.000000 \n",
1760
+ "75% 0.104000 136.045000 4.000000 6.000000 \n",
1761
+ "max 0.969000 244.035000 5.000000 655.000000 \n",
1762
+ "\n",
1763
+ " valence \n",
1764
+ "count 101939.000000 \n",
1765
+ "mean 0.482813 \n",
1766
+ "std 0.261690 \n",
1767
+ "min 0.000000 \n",
1768
+ "25% 0.271000 \n",
1769
+ "50% 0.477000 \n",
1770
+ "75% 0.693000 \n",
1771
+ "max 0.993000 "
1772
+ ],
1773
+ "text/html": [
1774
+ "\n",
1775
+ " <div id=\"df-a3275db4-0443-4f89-89b9-8ec83d22feda\">\n",
1776
+ " <div class=\"colab-df-container\">\n",
1777
+ " <div>\n",
1778
+ "<style scoped>\n",
1779
+ " .dataframe tbody tr th:only-of-type {\n",
1780
+ " vertical-align: middle;\n",
1781
+ " }\n",
1782
+ "\n",
1783
+ " .dataframe tbody tr th {\n",
1784
+ " vertical-align: top;\n",
1785
+ " }\n",
1786
+ "\n",
1787
+ " .dataframe thead th {\n",
1788
+ " text-align: right;\n",
1789
+ " }\n",
1790
+ "</style>\n",
1791
+ "<table border=\"1\" class=\"dataframe\">\n",
1792
+ " <thead>\n",
1793
+ " <tr style=\"text-align: right;\">\n",
1794
+ " <th></th>\n",
1795
+ " <th>Unnamed: 0</th>\n",
1796
+ " <th>acousticness</th>\n",
1797
+ " <th>danceability</th>\n",
1798
+ " <th>disc_number</th>\n",
1799
+ " <th>duration_ms</th>\n",
1800
+ " <th>energy</th>\n",
1801
+ " <th>instrumentalness</th>\n",
1802
+ " <th>key</th>\n",
1803
+ " <th>liveness</th>\n",
1804
+ " <th>loudness</th>\n",
1805
+ " <th>mode</th>\n",
1806
+ " <th>popularity</th>\n",
1807
+ " <th>speechiness</th>\n",
1808
+ " <th>tempo</th>\n",
1809
+ " <th>time_signature</th>\n",
1810
+ " <th>track_number</th>\n",
1811
+ " <th>valence</th>\n",
1812
+ " </tr>\n",
1813
+ " </thead>\n",
1814
+ " <tbody>\n",
1815
+ " <tr>\n",
1816
+ " <th>count</th>\n",
1817
+ " <td>101939.000000</td>\n",
1818
+ " <td>101939.000000</td>\n",
1819
+ " <td>101939.000000</td>\n",
1820
+ " <td>101939.000000</td>\n",
1821
+ " <td>1.019390e+05</td>\n",
1822
+ " <td>101939.000000</td>\n",
1823
+ " <td>101939.000000</td>\n",
1824
+ " <td>101939.000000</td>\n",
1825
+ " <td>101939.000000</td>\n",
1826
+ " <td>101939.000000</td>\n",
1827
+ " <td>101939.000000</td>\n",
1828
+ " <td>101939.000000</td>\n",
1829
+ " <td>101939.000000</td>\n",
1830
+ " <td>101939.000000</td>\n",
1831
+ " <td>101939.000000</td>\n",
1832
+ " <td>101939.000000</td>\n",
1833
+ " <td>101939.000000</td>\n",
1834
+ " </tr>\n",
1835
+ " <tr>\n",
1836
+ " <th>mean</th>\n",
1837
+ " <td>50969.000000</td>\n",
1838
+ " <td>0.352124</td>\n",
1839
+ " <td>0.586015</td>\n",
1840
+ " <td>1.032166</td>\n",
1841
+ " <td>2.467708e+05</td>\n",
1842
+ " <td>0.586479</td>\n",
1843
+ " <td>0.148776</td>\n",
1844
+ " <td>5.270858</td>\n",
1845
+ " <td>0.197640</td>\n",
1846
+ " <td>-9.462720</td>\n",
1847
+ " <td>0.618154</td>\n",
1848
+ " <td>39.782311</td>\n",
1849
+ " <td>0.128841</td>\n",
1850
+ " <td>118.358527</td>\n",
1851
+ " <td>3.875651</td>\n",
1852
+ " <td>4.608060</td>\n",
1853
+ " <td>0.482813</td>\n",
1854
+ " </tr>\n",
1855
+ " <tr>\n",
1856
+ " <th>std</th>\n",
1857
+ " <td>29427.398883</td>\n",
1858
+ " <td>0.334855</td>\n",
1859
+ " <td>0.177724</td>\n",
1860
+ " <td>0.566789</td>\n",
1861
+ " <td>1.904303e+05</td>\n",
1862
+ " <td>0.260170</td>\n",
1863
+ " <td>0.304024</td>\n",
1864
+ " <td>3.577679</td>\n",
1865
+ " <td>0.175391</td>\n",
1866
+ " <td>6.198508</td>\n",
1867
+ " <td>0.485841</td>\n",
1868
+ " <td>16.790769</td>\n",
1869
+ " <td>0.203324</td>\n",
1870
+ " <td>30.224074</td>\n",
1871
+ " <td>0.517008</td>\n",
1872
+ " <td>7.181805</td>\n",
1873
+ " <td>0.261690</td>\n",
1874
+ " </tr>\n",
1875
+ " <tr>\n",
1876
+ " <th>min</th>\n",
1877
+ " <td>0.000000</td>\n",
1878
+ " <td>0.000000</td>\n",
1879
+ " <td>0.000000</td>\n",
1880
+ " <td>1.000000</td>\n",
1881
+ " <td>1.155000e+03</td>\n",
1882
+ " <td>0.000000</td>\n",
1883
+ " <td>0.000000</td>\n",
1884
+ " <td>0.000000</td>\n",
1885
+ " <td>0.000000</td>\n",
1886
+ " <td>-60.000000</td>\n",
1887
+ " <td>0.000000</td>\n",
1888
+ " <td>0.000000</td>\n",
1889
+ " <td>0.000000</td>\n",
1890
+ " <td>0.000000</td>\n",
1891
+ " <td>0.000000</td>\n",
1892
+ " <td>1.000000</td>\n",
1893
+ " <td>0.000000</td>\n",
1894
+ " </tr>\n",
1895
+ " <tr>\n",
1896
+ " <th>25%</th>\n",
1897
+ " <td>25484.500000</td>\n",
1898
+ " <td>0.040700</td>\n",
1899
+ " <td>0.480000</td>\n",
1900
+ " <td>1.000000</td>\n",
1901
+ " <td>1.840000e+05</td>\n",
1902
+ " <td>0.411000</td>\n",
1903
+ " <td>0.000000</td>\n",
1904
+ " <td>2.000000</td>\n",
1905
+ " <td>0.095600</td>\n",
1906
+ " <td>-11.149000</td>\n",
1907
+ " <td>0.000000</td>\n",
1908
+ " <td>29.000000</td>\n",
1909
+ " <td>0.036400</td>\n",
1910
+ " <td>95.973000</td>\n",
1911
+ " <td>4.000000</td>\n",
1912
+ " <td>1.000000</td>\n",
1913
+ " <td>0.271000</td>\n",
1914
+ " </tr>\n",
1915
+ " <tr>\n",
1916
+ " <th>50%</th>\n",
1917
+ " <td>50969.000000</td>\n",
1918
+ " <td>0.238000</td>\n",
1919
+ " <td>0.610000</td>\n",
1920
+ " <td>1.000000</td>\n",
1921
+ " <td>2.168930e+05</td>\n",
1922
+ " <td>0.629000</td>\n",
1923
+ " <td>0.000037</td>\n",
1924
+ " <td>5.000000</td>\n",
1925
+ " <td>0.124000</td>\n",
1926
+ " <td>-7.599000</td>\n",
1927
+ " <td>1.000000</td>\n",
1928
+ " <td>41.000000</td>\n",
1929
+ " <td>0.050600</td>\n",
1930
+ " <td>118.067000</td>\n",
1931
+ " <td>4.000000</td>\n",
1932
+ " <td>2.000000</td>\n",
1933
+ " <td>0.477000</td>\n",
1934
+ " </tr>\n",
1935
+ " <tr>\n",
1936
+ " <th>75%</th>\n",
1937
+ " <td>76453.500000</td>\n",
1938
+ " <td>0.645000</td>\n",
1939
+ " <td>0.714000</td>\n",
1940
+ " <td>1.000000</td>\n",
1941
+ " <td>2.610550e+05</td>\n",
1942
+ " <td>0.798000</td>\n",
1943
+ " <td>0.034400</td>\n",
1944
+ " <td>8.000000</td>\n",
1945
+ " <td>0.241000</td>\n",
1946
+ " <td>-5.509000</td>\n",
1947
+ " <td>1.000000</td>\n",
1948
+ " <td>52.000000</td>\n",
1949
+ " <td>0.104000</td>\n",
1950
+ " <td>136.045000</td>\n",
1951
+ " <td>4.000000</td>\n",
1952
+ " <td>6.000000</td>\n",
1953
+ " <td>0.693000</td>\n",
1954
+ " </tr>\n",
1955
+ " <tr>\n",
1956
+ " <th>max</th>\n",
1957
+ " <td>101938.000000</td>\n",
1958
+ " <td>0.996000</td>\n",
1959
+ " <td>0.989000</td>\n",
1960
+ " <td>81.000000</td>\n",
1961
+ " <td>5.505831e+06</td>\n",
1962
+ " <td>1.000000</td>\n",
1963
+ " <td>1.000000</td>\n",
1964
+ " <td>11.000000</td>\n",
1965
+ " <td>0.999000</td>\n",
1966
+ " <td>2.719000</td>\n",
1967
+ " <td>1.000000</td>\n",
1968
+ " <td>97.000000</td>\n",
1969
+ " <td>0.969000</td>\n",
1970
+ " <td>244.035000</td>\n",
1971
+ " <td>5.000000</td>\n",
1972
+ " <td>655.000000</td>\n",
1973
+ " <td>0.993000</td>\n",
1974
+ " </tr>\n",
1975
+ " </tbody>\n",
1976
+ "</table>\n",
1977
+ "</div>\n",
1978
+ " <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a3275db4-0443-4f89-89b9-8ec83d22feda')\"\n",
1979
+ " title=\"Convert this dataframe to an interactive table.\"\n",
1980
+ " style=\"display:none;\">\n",
1981
+ " \n",
1982
+ " <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
1983
+ " width=\"24px\">\n",
1984
+ " <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
1985
+ " <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
1986
+ " </svg>\n",
1987
+ " </button>\n",
1988
+ " \n",
1989
+ " <style>\n",
1990
+ " .colab-df-container {\n",
1991
+ " display:flex;\n",
1992
+ " flex-wrap:wrap;\n",
1993
+ " gap: 12px;\n",
1994
+ " }\n",
1995
+ "\n",
1996
+ " .colab-df-convert {\n",
1997
+ " background-color: #E8F0FE;\n",
1998
+ " border: none;\n",
1999
+ " border-radius: 50%;\n",
2000
+ " cursor: pointer;\n",
2001
+ " display: none;\n",
2002
+ " fill: #1967D2;\n",
2003
+ " height: 32px;\n",
2004
+ " padding: 0 0 0 0;\n",
2005
+ " width: 32px;\n",
2006
+ " }\n",
2007
+ "\n",
2008
+ " .colab-df-convert:hover {\n",
2009
+ " background-color: #E2EBFA;\n",
2010
+ " box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
2011
+ " fill: #174EA6;\n",
2012
+ " }\n",
2013
+ "\n",
2014
+ " [theme=dark] .colab-df-convert {\n",
2015
+ " background-color: #3B4455;\n",
2016
+ " fill: #D2E3FC;\n",
2017
+ " }\n",
2018
+ "\n",
2019
+ " [theme=dark] .colab-df-convert:hover {\n",
2020
+ " background-color: #434B5C;\n",
2021
+ " box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
2022
+ " filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
2023
+ " fill: #FFFFFF;\n",
2024
+ " }\n",
2025
+ " </style>\n",
2026
+ "\n",
2027
+ " <script>\n",
2028
+ " const buttonEl =\n",
2029
+ " document.querySelector('#df-a3275db4-0443-4f89-89b9-8ec83d22feda button.colab-df-convert');\n",
2030
+ " buttonEl.style.display =\n",
2031
+ " google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
2032
+ "\n",
2033
+ " async function convertToInteractive(key) {\n",
2034
+ " const element = document.querySelector('#df-a3275db4-0443-4f89-89b9-8ec83d22feda');\n",
2035
+ " const dataTable =\n",
2036
+ " await google.colab.kernel.invokeFunction('convertToInteractive',\n",
2037
+ " [key], {});\n",
2038
+ " if (!dataTable) return;\n",
2039
+ "\n",
2040
+ " const docLinkHtml = 'Like what you see? Visit the ' +\n",
2041
+ " '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
2042
+ " + ' to learn more about interactive tables.';\n",
2043
+ " element.innerHTML = '';\n",
2044
+ " dataTable['output_type'] = 'display_data';\n",
2045
+ " await google.colab.output.renderOutput(dataTable, element);\n",
2046
+ " const docLink = document.createElement('div');\n",
2047
+ " docLink.innerHTML = docLinkHtml;\n",
2048
+ " element.appendChild(docLink);\n",
2049
+ " }\n",
2050
+ " </script>\n",
2051
+ " </div>\n",
2052
+ " </div>\n",
2053
+ " "
2054
+ ]
2055
+ },
2056
+ "metadata": {},
2057
+ "execution_count": 22
2058
+ }
2059
+ ]
2060
+ }
2061
+ ]
2062
+ }
requirements.txt ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ streamlit==1.0.0
2
+ pandas==1.3.3
3
+ plotly==5.3.1
4
+ scikit-learn==0.23.2