# -*- coding: utf-8 -*- """Most Popular Albums Per Artist With Gradio.ipynb Automatically generated by Colaboratory. Original file is located at https://colab.research.google.com/drive/1wMpev3zhNcOdO_amUdfeE6HCflOB8KIL # Set up Spotify credentials Before getting started you need: * Spotify API permissions & credentials that could apply for [here](https://developer.spotify.com/). Simply log in, go to your “dashboard” and select “create client id” and follow the instructions. Spotify are not too strict on providing permissions so put anything you like when they ask for commercial application. * Python module — spotipy — imported """ import spotipy #To access authorised Spotify data - https://developer.spotify.com/ from spotipy.oauth2 import SpotifyClientCredentials from fuzzywuzzy import fuzz import pandas as pd import seaborn as sns import gradio as gr #https://gradio.app/docs/#i_slider import matplotlib.pyplot as plt import time import numpy as np #Create Function for identifying your artist def choose_artist(name_input, sp): results = sp.search(name_input) #result_1 = result['tracks']['items'][0]['artists'] top_matches = [] counter = 0 #for each result item (max 10 I think) for i in results['tracks']['items']: #store current item current_item = results['tracks']['items'][counter]['artists'] counter+=1 #for each item in that search_term counter2 = 0 for i in current_item: #append artist name to top_matches #I will need to append something to identify the correct match, please update once I know top_matches.append((current_item[counter2]['name'], current_item[counter2]['uri'])) counter2+=1 #remove duplicates by turning list into a set, then back into a list top_matches = list(set(top_matches)) fuzzy_matches = [] #normal list doesn't need len(range) for i in top_matches: #put ratio result in variable to avoid errors ratio = fuzz.ratio(name_input, i[0]) #store as tuple but will need to increase to 3 to include uid fuzzy_matches.append((i[0], ratio, i[1])) #sort fuzzy matches by ratio score fuzzy_matches = sorted(fuzzy_matches, key=lambda tup: tup[1], reverse=True) #store highest tuple's attributes in chosen variables chosen = fuzzy_matches[0][0] chosen_id = fuzzy_matches[0][1] chosen_uri = fuzzy_matches[0][2] print("The results are based on the artist: ", chosen) return chosen, chosen_id, chosen_uri #Function to Pull all of your artist's albums def find_albums(artist_uri, sp): sp_albums = sp.artist_albums(artist_uri, album_type='album', limit=50) #There's a 50 album limit album_names = [] album_uris = [] for i in range(len(sp_albums['items'])): #Keep names and uris in same order to keep track of duplicate albums album_names.append(sp_albums['items'][i]['name']) album_uris.append(sp_albums['items'][i]['uri']) return album_uris, album_names #Function to store all album details along with their song details def albumSongs(album, sp, album_count, album_names, spotify_albums): spotify_albums[album] = {} #Creates dictionary for that specific album #Create keys-values of empty lists inside nested dictionary for album spotify_albums[album]['album_name'] = [] #create empty list spotify_albums[album]['track_number'] = [] spotify_albums[album]['song_id'] = [] spotify_albums[album]['song_name'] = [] spotify_albums[album]['song_uri'] = [] tracks = sp.album_tracks(album) #pull data on album tracks for n in range(len(tracks['items'])): #for each song track spotify_albums[album]['album_name'].append(album_names[album_count]) #append album name tracked via album_count spotify_albums[album]['track_number'].append(tracks['items'][n]['track_number']) spotify_albums[album]['song_id'].append(tracks['items'][n]['id']) spotify_albums[album]['song_name'].append(tracks['items'][n]['name']) spotify_albums[album]['song_uri'].append(tracks['items'][n]['uri']) #Add popularity category def popularity(album, sp, spotify_albums): #Add new key-values to store audio features spotify_albums[album]['popularity'] = [] #create a track counter track_count = 0 for track in spotify_albums[album]['song_uri']: #pull audio features per track pop = sp.track(track) spotify_albums[album]['popularity'].append(pop['popularity']) track_count+=1 def gradio_music_graph(client_id, client_secret, artist_name): #total_albums #Insert your credentials client_credentials_manager = SpotifyClientCredentials(client_id=client_id, client_secret=client_secret) sp = spotipy.Spotify(client_credentials_manager=client_credentials_manager) #spotify object to access API #Choose your artist via input chosen_artist, ratio_score, artist_uri = choose_artist(artist_name, sp=sp) #Retrieve their album details album_uris, album_names = find_albums(artist_uri, sp=sp) #Create dictionary to store all the albums spotify_albums = {} #Album count tracker album_count = 0 for i in album_uris: #for each album albumSongs(i, sp=sp, album_count=album_count, album_names=album_names, spotify_albums=spotify_albums) print("Songs from " + str(album_names[album_count]) + " have been added to spotify_albums dictionary") album_count+=1 #Updates album count once all tracks have been added #To avoid it timing out sleep_min = 2 sleep_max = 5 start_time = time.time() request_count = 0 #Update albums with popularity scores for album in spotify_albums: popularity(album, sp=sp, spotify_albums=spotify_albums) request_count+=1 if request_count % 5 == 0: # print(str(request_count) + " playlists completed") time.sleep(np.random.uniform(sleep_min, sleep_max)) # print('Loop #: {}'.format(request_count)) # print('Elapsed Time: {} seconds'.format(time.time() - start_time)) #Create song dictonary to convert into Dataframe dic_df = {} dic_df['album_name'] = [] dic_df['track_number'] = [] dic_df['song_id'] = [] dic_df['song_name'] = [] dic_df['song_uri'] = [] dic_df['popularity'] = [] for album in spotify_albums: for feature in spotify_albums[album]: dic_df[feature].extend(spotify_albums[album][feature]) #Convert into dataframe df = pd.DataFrame.from_dict(dic_df) df = df.sort_values(by='popularity') df = df.drop_duplicates(subset=['song_id'], keep=False) sns.set_style('ticks') fig, ax = plt.subplots() fig.set_size_inches(11, 8) ax.set_xticklabels(ax.get_xticklabels(), rotation=25, ha="right", wrap=True) plt.tight_layout(rect=[1, 2.5, 1, 0.45]) #(left, bottom, right, top) sns.boxplot(x=df["album_name"], y=df["popularity"], ax=ax) plt.show() return fig #Interface will include these buttons based on parameters in the function with a dataframe output music_plots = gr.Interface(gradio_music_graph, ["text", "text", "text"], ["plot"], title="Popular Songs By Album Box Plot Distribution on Spotify", description="Using your Spotify API Access from https://developer.spotify.com/ you can see your favourite artist's most popular albums on Spotify") music_plots.launch(debug=True)