webapp.py

from tkinter.font import names
from click import option
from dash import *
#from matplotlib.pyplot import figure
import plotly.express as px
import plotly.graph_objects as go
import pandas as pd
from pyrsistent import v
import processing as pp
import emotions as em

user_df = pd.DataFrame()
playlists = []
songsList = []

app = Dash(__name__)
app.layout = html.Div(className='row',children=[
    html.H1(children='Analizador de canciones'),

    html.Div(children='''
        Analiza la canción y los sentimientos que transmiten la música que escuchas en Spotify
    '''),
    html.Br(),
    html.Div(children=[
        html.Label(children='Introduce tu usuario '),
        dcc.Input(
            id="inputuser",
            type='text',
            placeholder="Escribe aquí el usuario",
        ),
        html.Button(id='buscar',children='Buscar',n_clicks=0)
    ]),
    html.Br(),
    html.Div(id='sel-playlist',children=[
        html.Label(children='Selecciona una playlist '),
        dcc.Dropdown(id='playlist'),
    ]),
    
    html.Br(),
    html.Div(children=[
        html.Div(children=[
            html.Label(children='Características de canciones de la playlist'),
            dcc.Dropdown(id='song1'),
            html.Button(id='reset-polar',children='Reset',n_clicks=0),
            dcc.Graph(id='polar',figure=go.Figure())
        ],className="row",style={'width': '49%','display': 'inline-block'}),
        html.Div(children=[
            html.Label(children='Gráfica de canciones explicitas de la playlist seleccionda'),
            html.Br(),
            dcc.Graph(id='explicit-graph',figure=go.Figure())
        ],className="row",style={'width': '49%','display': 'inline-block'}),
    ]),
    html.Br(),
    html.Div(children=[
        html.Br(),
            html.Label(children='Indica el número de canciones a analizar'),
            dcc.Slider(id='slider',min=1,value=1,step=1,max=2),
            html.Button(id='analize',children='Analizar',n_clicks=0),
            html.Label(children='Gráfica de sentimiento de la playlist seleccionada'),
            html.Br(),
            dcc.Graph(id='sentiment-graph'),
    ]),
])

@app.callback(
    Output('playlist','options'),
    Input('buscar', 'n_clicks'),
    State('inputuser','value'),
)
def get_playlists(clicks,value):
    global user_df
    user_df = pd.DataFrame()
    user_df = pp.getDataframeOfUser(value)
    res = user_df['PlaylistName'].unique()
        

    #results = get_playlists_user(value)
    # df = pd.DataFrame()
    # for item in user_df['PlaylistName']:
    #     tracknames = item['name']
    #     ids = item['id']
    # df['name'] = tracknames
    # df['id'] = ids
    # print(df.head())
    # playlists = {}
    # playlists = {'label':results['items'][0]['name'],'value':results['items'][0]['id']} #0 no, todas

    return res

@app.callback(
    Output('slider','max'),
    Output('song1','options'),
    Output('song1','value'),
    Input('playlist','value'),
)
def get_songs(value):
    global user_df, songsList
    songs_df = user_df[user_df['PlaylistName']==value]
    songsList = []
    options = []
    for index, row in songs_df.iterrows():
        songsList.append({'artist':row['ArtistName'],'song':row['SongName']})
        options.append({'label':row['ArtistName'] + ' - ' + row['SongName'],'value':row['SongId']})
    
    return len(songsList), options, options[0]['value']
    
    
@app.callback(
    Output('sentiment-graph','figure'),
    Input('analize', 'n_clicks'),
    State('slider','value')
)
def show_graph(nclicks, value):
    global songsList,user_df
    songsList = songsList[0:value]
    print(songsList)
    res_df = em.getLyricsInfo(songsList)
    #merged_df = user_df.merge(res_df,'left',left_on='SongName',right_on='title')
    res_df.to_csv('carlos2.csv',index='false')

    #merged_df.to_csv('carlos3.csv',index='false')
    max_range = max(res_df['positive'].max(),res_df['negative'].max())
    max_range *= 1.05
    fig = []
    fig = px.scatter(res_df,'positive', 'negative',range_x=[-0.01,max_range], range_y=[-0.01,max_range],symbol='artist',color="title")
    return fig

@app.callback(
    Output('explicit-graph','figure'),
    Input('playlist','value')
)
def show_explicitness(value):
    global user_df
    songs = user_df[user_df['PlaylistName']==value]
    explicitness = songs['SongIsExplicit'].tolist()
    numYes = 0
    numNo = 0
    #Contamos las columnas de explicit
    for song in explicitness:
        if song == True:
            numYes = numYes + 1
        else:
            numNo = numNo + 1
    values = [numYes, numNo]
    names = ['Explicita', 'Familiar']
    
    fig = px.pie(values=values, names=names)
    return fig

# songs_df = user_df[user_df['PlaylistName']==value]
#     songsList = []
#     for row in songs_df.iterrows():
#         songsList.append({'explicit':row['SongIsExplicit']})
# @app.callback(
#     Output('polar','figure'),
#     Input('song1','value')
# )
# def show_polar(value):
#     data,tempo = pp.getAudioFeatures(value)
#     fig = px.line_polar(data, r='r', theta='theta', line_close=True)
#     fig.update_traces(fill='toself')
#     return fig

@app.callback(
    Output('polar','figure'),
    Input('playlist','valuePlay'),
    Input('song1','value'),

    Input('reset-polar', 'n_clicks'),
    State('polar','figure')
)
def show_polar(valuePlaylist,value,clicks, figure):
    ctx = callback_context

    if not ctx.triggered:
        button_id = 'No clicks yet'
    else:
        button_id = ctx.triggered[0]['prop_id'].split('.')[0]
        
    if(button_id=='reset-polar' or button_id=='playlist'):
        figure['data'] = []
        return figure
    else:
        data,tempo = pp.getAudioFeatures(value)
        #print(figure)
        if(figure['data']!=None):
            figure['data'].append(go.Scatterpolar(r=data['r'],
                theta=data['theta'],
                fill='toself',
                name=user_df.loc[user_df['SongId']==value, 'ArtistName'].iloc[0] + ' - ' + user_df.loc[user_df['SongId']==value, 'SongName'].iloc[0]
            ))
        # else:
        #     figure['data']=[go.Scatterpolar(r=data['r'],
        #         theta=data['theta'],
        #         fill='toself',
        #         name = user_df.loc[user_df['SongId']==value, 'SongName'].iloc[0]
        #         #name=user_df[user_df['SongId']==value]['ArtistName'].item() + ' - ' + user_df[user_df['SongId']==value]['SongName'].item()
        #     )]
        return figure


if __name__ == '__main__':
    app.run_server(debug=False)