I’m building a dashboard that displays a graph for one of two distributions selected in a dropdown. When a user selects the Beta distribution, the Multivariate Normal distribution graph is of course suppressed, and vice-versa.
The dashboard works, but it gives the
‘A nonexistent object was used in an Input of a Dash callback.’ error. Not sure what the correct and elegant way to resolve the error is.
from dash import Dash, html, dcc, Input, Output, State
import dash_bootstrap_components as dbc
import numpy as np
import plotly.graph_objects as go
import plotly.express as px
from scipy import stats
# App Layout *******************************************
app = Dash(__name__, external_stylesheets=[dbc.themes.LITERA])
app.config.suppress_callback_exceptions = True
app.layout = dbc.Container(
[
dbc.Row
(
dbc.Col(
html.H2(
"Visualizing Distributions",
className="text-center bg-primary text-white p-2",
),
)
),
dbc.Row(
[
dbc.Col(
[html.H3(
"Select distribution",
style={"textAlign": "center"},
),
dcc.Dropdown(
id='select-distribution',
options=[
{'label': 'Beta', 'value': 'beta'},
{'label': 'Multivariate normal', 'value': 'multivariate_normal'}
],
value='beta',
),
html.Div(
id = 'sliders-or-inputs',
className="mt-4 p-4"
)
],
width=4
),
dbc.Col(
[
html.Div([], id='beta_graph'),
html.Div([], id='mv_graph'),
],
width=8
),
]
)
]
)
# Callbacks *******************************************
@app.callback(
Output('sliders-or-inputs', 'children'),
[Input('select-distribution', 'value')])
def display_sliders_or_inputs(distribution):
if distribution == 'beta':
return html.Div([
dcc.Slider(
id='alpha',
min=0,
max=100,
value=0,
className='four columns',
tooltip={"placement": "bottom", "always_visible": True}
),
dcc.Slider(
id='beta',
min=0,
max=100,
value=0,
className='four columns',
tooltip={"placement": "bottom", "always_visible": True}
)
], className='row')
elif distribution == 'multivariate_normal':
return html.Div([
dbc.InputGroup(
[
dbc.InputGroupText("1st variable mean"),
dbc.Input(
id="var1_mu",
placeholder="10",
type="number",
min=10,
value=58.8,
),
],
className="mb-3",
),
dbc.InputGroup(
[
dbc.InputGroupText("1st variable standard deviation"),
dbc.Input(
id="var1_sd",
placeholder="3",
type="number",
min=0.01,
value=7.82,
),
],
className="mb-3",
),
dbc.InputGroup(
[
dbc.InputGroupText("2nd variable mean"),
dbc.Input(
id="var2_mu",
placeholder="7",
type="number",
min=10,
value=60.7,
),
],
className="mb-3",
),
dbc.InputGroup(
[
dbc.InputGroupText("2nd variable standard deviation"),
dbc.Input(
id="var2_sd",
placeholder="10",
type="number",
min=0.01,
value=7.6,
),
],
className="mb-3",
),
dbc.InputGroup(
[
dbc.InputGroupText("Variable correlation"),
dbc.Input(
id="corr",
placeholder="0.12",
type="number",
min=-1,
value=0.7,
),
],
className="mb-3",
),
], className='row')
@app.callback(
Output("mv_graph", "children"),
Input('select-distribution', 'value'),
Input("var1_mu", "value"),
Input("var1_sd", "value"),
Input("var2_mu", "value"),
Input("var2_sd", "value"),
Input("corr", "value"),
)
def update_mv_normal_distr(distribution, var1_mu, var1_sd, var2_mu, var2_sd, corr):
if distribution=='multivariate_normal':
x1, x2 = np.mgrid[40:80:0.25, 40:80:0.25]
means = np.array([var1_mu, var2_mu])
cov = var1_sd * var2_sd * corr
cov_mat = np.array([[var1_sd**2, cov], [cov, var2_sd**2]])
z = stats.multivariate_normal(means, cov_mat).pdf(np.dstack((x1, x2)))
fig = go.Figure(data=[go.Surface(z=z)])
fig.update_xaxes(title_text='Testing distribution display')
fig.update_yaxes(title_text='Probability Density')
fig.update_layout(width=700, height=700, title="Multivariate Normal Distribution",
scene = dict(xaxis = dict(title = 'Var1'),
yaxis = dict(title = 'Var2'),
zaxis = dict(title = 'Probability density')),
margin=dict(l=0, r=50, b=50, t=50))
return dcc.Graph(figure=fig)
else:
return None
@app.callback(
Output("beta_graph", "children"),
Input('select-distribution', 'value'),
Input("alpha", "value"),
Input("beta", "value")
)
def update_beta_distr(distribution, a, b):
if distribution=='beta':
x = np.linspace(0, 1, 10000)
beta_distr = stats.beta(a,b).pdf(x)
fig = px.histogram(beta_distr, title="Beta distribution")
return dcc.Graph(figure=fig)
else:
return None
if __name__ == "__main__":
app.run_server(debug=True)
I found this SO post. The solution feels more complex than necessary, but I’m new to Dash.
Also, general critiques re: app architecture as well as formatting welcome, since I’m learning.
