/
music_data.py
151 lines (109 loc) · 5.21 KB
/
music_data.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
# Getting data
import requests # to access the APIs
import json
import unicodecsv as csv
from pprint import PrettyPrinter
import spotipy
from spotipy.oauth2 import SpotifyClientCredentials
# Analyzing data
import pandas as pd
import numpy as np
import scipy
from scipy.stats import ttest_ind
from pathlib import Path
import secret # get API keys
# Definitions
LF_BASE = 'https://ws.audioscrobbler.com/2.0/'
# Authenticate spotify
client_credentials_manager = SpotifyClientCredentials(client_id=secret.SPOTIPY_CLIENT_ID, client_secret=secret.SPOTIPY_CLIENT_SECRET)
sp = spotipy.Spotify(client_credentials_manager=client_credentials_manager)
features = ['acousticness', 'danceability', 'energy', 'speechiness', 'tempo', 'valence']
pp = PrettyPrinter(indent=2)
# Returns an array of similar artists using Last.fm API
def get_similar_artists(artist, num_artists=10):
similar_artists_params = {'method': 'artist.getSimilar', 'artist': artist, 'limit': num_artists, 'api_key': secret.LF_KEY, 'format': 'json'}
similar_artists_request = requests.get(LF_BASE, params=similar_artists_params).json()
similar_artists = []
# pp.pprint(similar_artists_request["similarartists"]["artist"])
for artist in similar_artists_request["similarartists"]["artist"]:
similar_artists.append(artist["name"])
return similar_artists
def get_genre_top_artists(genre, num_artists=10):
# Get a genre's top artists
top_artists_params = {'method': 'tag.gettopartists', 'tag': genre, 'limit': num_artists, 'api_key': secret.LF_KEY, 'format': 'json'}
top_artists_request = requests.get(LF_BASE, params=top_artists_params).json()
top_artists = []
for artist in top_artists_request["topartists"]["artist"]:
top_artists.append(artist["name"])
return top_artists
# Writes artist's top tracks to the CSV defined by the writer
def write_artist_top_tracks_data(writer, artist):
# Get data about the artist
artist_data_params = {'method': 'artist.gettoptracks', 'artist': artist, 'limit': 10, 'api_key': secret.LF_KEY, 'format': 'json'}
artist_data_request = requests.get(LF_BASE, params=artist_data_params).json()
# For each track, get music data from spotify
for track in artist_data_request['toptracks']['track']:
song_name = track['name']
audio_features = get_music_features(artist, song_name)
if len(audio_features) > 0:
write_to_csv(writer, artist, song_name, audio_features)
# Given a song and arist name, find that song on Spotify and return
# an array of audio features for that song
def get_music_features(artist, song_name):
q = 'artist:' + artist + " track:" + song_name
sp_search = sp.search(q, type="track", limit=1)
if len(sp_search['tracks']['items']) >= 1:
song_id = sp_search['tracks']['items'][0]['id']
# Get audio features
audio_features = sp.audio_features(tracks=[song_id])
song_features = []
for feature in features:
song_features.append(audio_features[0][feature])
return song_features
return []
# Given a CSV writer, artist name, song name, and an array of audio features,
# write that data to a CSV file
def write_to_csv(writer, artist, song_name, audio_features):
writer.writerow([artist, song_name] + audio_features)
def get_data(csv_name, artists):
# Set up CSV file for pandas
f = open(csv_name, 'wb')
writer = csv.writer(f, delimiter="|", quotechar="", quoting=csv.QUOTE_NONE, encoding='utf-8')
writer.writerow(['artist', 'song'] + features)
# Get data about all artists
for artist in artists:
write_artist_top_tracks_data(writer, artist)
f.close()
def compare_two_artists(artist1_name, artist1_df, artist2_name, artist2_df, features, file):
num_diffs = 0
for feature in features:
# Print means
artist1_avg = artist1_df[feature].mean()
file.write(("Average {0} for {1} is {2}\n").format(feature, artist1_name, str(artist1_avg)))
artist_avg = artist2_df[feature].mean()
file.write(("Average {0} for {1} is {2}\n").format(feature, artist2_name, str(artist_avg)))
# Do T-Test
(stat, pvalue) = ttest_ind(artist1_df[feature], artist2_df[feature])
if pvalue < 0.05:
file.write(("P value is {0}, so there is NO statistically significant difference\n\n").format(pvalue))
else:
num_diffs += 1
file.write(("P value is {0}, so there IS a statistically significant difference\n\n").format(pvalue))
file.write(("{0} and {1} differ in {2} out of {3} categories\n\n\n").format(artist1_name, artist2_name, num_diffs, len(features)))
# Analyzes data in csv_name, so assumes data has already been retrieved
# Compares the artists in group1 with the artists in group2
def analyze_data(csv_name, group1, group2, file_name):
# Make sure that the csv file exists
if not Path(csv_name).is_file():
print("Cannot access {0}", csv_name)
return
# Read data into pandas dataframe
music_data = pd.read_csv(csv_name, sep='|', error_bad_lines=False)
f = open(file_name, 'w')
for artist1 in group1:
for artist2 in group2:
f.write("Comparing " + artist1 + " and " + artist2 + "\n\n")
artist1_df = music_data.query(('artist == "{}"').format(artist1))
artist2_df = music_data.query(('artist == "{}"').format(artist2))
compare_two_artists(artist1, artist1_df, artist2, artist2_df, features, f)
f.close()