✊🏿 Black Lives Matter. Please consider donating to Black Girls Code today.
⚡️ Concerned about the grid? Kyle Baranko teaches how to predicting peak loads using XGBoost. Register for the August webinar!

Calling Plotly.plot as if redrawing but this container doesn't yet have a plot

Hi everyone, I’ve made another dashboard for Covid19 in Italy, but with region selectable with a dropdown.
For the styling I’m using dash-bootstrap-components, but my dashboard layout behaves in a strange way.
If I open my browser console, I get the error of the title of this post, and in debug mode I get this other error “Warning: Cannot update a component from inside the function body of a different component.”.
Practically through a callback associated with a function for each chart, I update the data based on the region selected in the dropdown.
But in some regions, the plots are missing…
You can see my test dashboard here.

And I post also two pieces of code:

dbc.Row(  # andamento contagi
            dbc.Col(
                dcc.Graph(
                    id='andamento-contagi',
                    figure={},
                    config=chart_config
                )
                , width=12)
        ),

and the relative callback:

@app.callback(
    Output('andamento-contagi', 'figure'),
    [Input('region_select', 'value')])
def update_andamento_contagi(regione):
    reg_df = df.loc[df['denominazione_regione'] == regione]
    local_df = calculate_data(reg_df.copy())
    figure = {
        'data': [
            {'x': local_df['data'], 'y': local_df['nuovi_positivi'], 'type': 'bar', 'name': 'Nuovi Casi'},
            {'x': local_df['data'], 'y': local_df['nuovi_positivi_avg'], 'type': 'scatter',
             'line': dict(color='orange'),
             'name': 'Media 7 giorni'}
        ],
        'layout': {
            'title': 'Andamento dei contagi',
            'xaxis': dict(
                rangeselector=dict(buttons=slider_button),
                rangeslider=dict(visible=False),
                type='date'
            )
        }
    }
    return figure

The code for the other plots is similar.