Error : UnboundLocalError: local variable 'fig' referenced before assignment

Dash Python
1

I’m having this error on the dash server, I believe this error is linked to the callback function
I’ve looked in several places and I don’t know how to solve it, you are my last chance :frowning:

opcoes = list(df['Colunas'].unique())

app.layout = html.Div(children=[

        html.H5('Opcões'),
        dcc.Dropdown(opcoes, value = 'ACS' , id='check_opcao'),

        dcc.Graph(figure=fig, id='grafico_gc_players')
        
        
        ])
@app.callback([
        Output('grafico_gc_players', 'figure'),
        
        ], 
                [
                Input('check_opcao', 'value'),
                
                ])

def renderizar_graficos(check_opcao):


    if check_opcao == 'ACS':
        fig_acs = px.bar(df, x='Player', y=check_opcao,  color='Player', barmode = 'stack', 
            labels={
                    'Player': 'Nome e Time'}, template='plotly_dark')
    else:
        tabela_filtrada = df.loc[df['Colunas'] == check_opcao,:]
        fig_geral = px.bar(tabela_filtrada, x='Player', y=check_opcao, color='Player', barmode = 'stack', 
            labels={
                    'Player': 'Nome e Time'}, template='plotly_dark')

    return fig_acs, fig_geral
2

@Luisanches,

Where you are putting dcc.graph in the app.layout, remove the statement

figure=fig

This should allow your callback to fire and fill in the figure as desired.

3

I removed the instruction but the error remains the same

UnboundLocalError: local variable 'fig_geral' referenced before assignment
4

Change the function to this:

def renderizar_graficos(check_opcao):


    if check_opcao == 'ACS':
        fig = px.bar(df, x='Player', y=check_opcao,  color='Player', barmode = 'stack', 
            labels={
                    'Player': 'Nome e Time'}, template='plotly_dark')
    else:
        tabela_filtrada = df.loc[df['Colunas'] == check_opcao,:]
        fig = px.bar(tabela_filtrada, x='Player', y=check_opcao, color='Player', barmode = 'stack', 
            labels={
                    'Player': 'Nome e Time'}, template='plotly_dark')

    return fig
5

I did that and the graph appeared again, but the callback doesn’t work and another error appeared…
I think the error is related to the way you built this function… but I don’t know how to identify the error

 Expected type: (<class 'tuple'>, <class 'list'>)
                Received value of type <class 'plotly.graph_objs._figure.Figure'>:
                    Figure({
    'data': [{'alignmentgroup': 'True',
              'hovertemplate': 'Nome e Time=%{y}<extra></extra>',
              'legendgroup': 'Bizinha  MIBR.F',
              'marker': {'color': '#636efa', 'pattern': {'shape': ''}},
              'name': 'Bizinha  MIBR.F',
              'offsetgroup': 'Bizinha  MIBR.F',
              'orientation': 'v',
              'showlegend': True,
              'textposition': 'auto',
              'type': 'bar',
              'x': array(['Bizinha  MIBR.F'], dtype=object),
              'xaxis': 'x',
              'y': array(['Bizinha  MIBR.F'], dtype=object),
              'yaxis': 'y'}],
    'layout': {'barmode': 'stack',
               'legend': {'title': {'text': 'Nome e Time'}, 'tracegroupgap': 0},
               'margin': {'t': 60},
               'template': '...',
               'xaxis': {'anchor': 'y',
                         'categoryarray': [Bizinha  MIBR.F],
                         'categoryorder': 'array',
                         'domain': [0.0, 1.0],
                         'title': {'text': 'Nome e Time'}},
               'yaxis': {'anchor': 'x',
                         'categoryarray': [Bizinha  MIBR.F],
                         'categoryorder': 'array',
                         'domain': [0.0, 1.0],
                         'title': {'text': 'Nome e Time'}}}
})
6

Try this. It looks like the issue is with you listing the outputs in the callback, ie ([Output(‘grafico_gc_players’, ‘figure’),]) I just removed this bit since you are just targeting one figure.

opcoes = list(df['Colunas'].unique())

app.layout = html.Div(children=[

        html.H5('Opcões'),
        dcc.Dropdown(opcoes, value = 'ACS' , id='check_opcao'),

        dcc.Graph(figure=fig, id='grafico_gc_players')
        
        
        ])
@app.callback(
        Output('grafico_gc_players', 'figure'),
                [
                Input('check_opcao', 'value'),
                
                ])

def renderizar_graficos(check_opcao):


    if check_opcao == 'ACS':
        fig = px.bar(df, x='Player', y=check_opcao,  color='Player', barmode = 'stack', 
            labels={
                    'Player': 'Nome e Time'}, template='plotly_dark')
    else:
        tabela_filtrada = df.loc[df['Colunas'] == check_opcao,:]
        fig = px.bar(tabela_filtrada, x='Player', y=check_opcao, color='Player', barmode = 'stack', 
            labels={
                    'Player': 'Nome e Time'}, template='plotly_dark')

    return fig
7

It worked! only that the graph has another problem, it is not showing the correct information. but this I’ll try to solve, if not I’ll open another topic, thanks for the help

8

I think you should change figure=fig to figure={}. I will return fig based on your dropdown.

9

Depends on if he already has a pre-defined figure that was passed as fig, but yes. You are right.