Customdata missing in hoverdata in Ploly 6.0.0

I have written an app in Python using Dash 2.18.2. In that app I have a map with a cycling track and several graphs to show distance vs. speed, distance vs. power etc. I have written an JS-script to show hover-info on the graphs when hovering over the map and vice versa. I use therefore 'customdata' that is encapsulated in the hoverdata. with this Dash code:

        customdata = np.stack((df0['Dist'], overall_focus, df0['Lons'], df0['Lats']), axis=1)
        hovertemplate=('Dist: %{customdata[0]:.1f}<br>'+
                      focus+': '+ '%{customdata[1]:.1f}<br>'+
                      '%{customdata[2]:.6f}<br>'+
                      '%{customdata[3]:.6f}<br>'+
                      '<extra></extra>')
       
    
        # Create the scatter map figure with plotly.express
        fig_trck = px.scatter_map(
            map_style=map_mode,
            center={'lat': center[1], 'lon': center[0]},
            zoom=zoom,
            lat=overall_lats,
            lon=overall_lons
        )
    
        # Update traces with custom data and styling
        fig_trck.update_traces(
            customdata=customdata,
            hovertemplate=hovertemplate,
            marker_size=3,
            marker_color=overall_focus,
            marker_colorscale='Rainbow',
            marker_showscale=False,
            line_width=2,
            line_color='Blue',
            mode=mode
        )

When I use Plotly 5.24.1 it gives this result in the DOM:
Object
points:Array(1)
0:
bbox:{x0: 894.0113118377704, x1: 896.0113118377704, y0: 354.0555201566335, y1: 356.0555201566335}
curveNumber: 0
customdata :( 4) [33519.2, 22.104, 4.507961, 51.843206]
lat:51.843206
lon:4.507961
pointIndex:4327
pointNumber:4327
[[Prototype]]:Object

when I use Plotly 6.0.0, I get this result:
Object
points : Array(1)
0:
bbox: {x0: 894.0113118377704, x1: 896.0113118377704, y0: 354.0555201566335, y1: 356.0555201566335}
curveNumber:0
customdata : undefined
lat:51.843206
lon:4.507961
pointIndex:4327
pointNumber:4327
[[Prototype]]: Object
length: 1

Can anyone explain this? Is it a bug or is customdata removed to another position in the DOM?

@Hermeij

@Hermeij I can't help you with this question technically ... but I AM super intrigued wrt what you're looking to display -> specifically "cycling track and several graphs to show distance vs. speed, distance vs. power etc." ... Do you have any screenshots, animated gifs, video of that??

@dgibbon-plotly

@dgibbon-plotly
Here is a link to a video I made for you:
https://1drv.ms/v/s!AuWRFk9aKH8Oi-QsAyQoccIL_PgWBA?e=WKaSVw

@gvwilson

Hi @gvwilson, @Hermeij – I've been experimenting with plotly6, and it seems that this issue occurs when customdata is not provided explicitly as a Python list. If customdata is passed as a numpy array or pandas Series instead (similar to how x and y trace properties are handled), the data gets base64-encoded, likely causing customdata to become undefined, hinting towards no consistent handling of encoded data across all trace-related properties!!.

When this is fixed, this should at least result in several tests (for other array-alike traces as well)!

(In my use-case, I tried to utilize the customdata of points during a clientside trace SelectedData callback)

📷 ⬇ ~ customdata when using list and plotly6.0.0

📷 ⬇ ~ customdata when using numpy array / pd.Series

@jonasvdd

@jonasvdd Thanks for your research into this issue. For now I will use an explicit list for customdata.

@Hermeij

@Hermeij Would you be able to provide a minimal code snippet which reproduces the issue you are seeing?

The code you provided gives a general idea, but the variables df0 and overall_focus are not defined within the code so I can't be sure what types they are.

I just merged a PR #7608 which might possibly resolve this issue and I'd like to verify.

@emilykl

@emilykl

This code gave at that time no customdata in hoverinfo. I see now that it is working correct with dash 3.2.0 and plotly 6.3.1

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Wed Oct 29 15:38:38 2025

@author: herman
"""

import dash
from dash import dcc, html
import pandas as pd
import numpy as np
import plotly.graph_objects as go

# Sample data
df0 = pd.DataFrame([[100.0, 51.0, 4.5],
                    [150.0, 51.1, 4.1]],
                   columns=['Dist', 'Lons', 'Lats'])

# Prepare customdata for hovertemplate
customdata = np.stack((df0['Dist'], df0['Lons'], df0['Lats']), axis=1)


# Define hovertemplate
hovertemplate = (
    'Dist: %{customdata[0]:.1f}<br>' +
    'Lon: %{customdata[1]:.1f}<br>' +
    'Lat: %{customdata[2]:.6f}<br>' +
    '<extra></extra>'
)

# Create figure
fig = go.Figure(go.Scattermap(
    mode='lines+markers',
    lon=df0['Lons'],
    lat=df0['Lats'],
    customdata=customdata,
    hovertemplate=hovertemplate,
    line=dict(width=2, color='blue'),
    marker=dict(size=8, color='red')
))

# Set mapbox style and center
fig.update_layout(
    map=dict(
        style='open-street-map',
        center=dict(lat=df0['Lats'].mean(), lon=df0['Lons'].mean()),
        zoom=5
    ),
    margin=dict(l=0, r=0, t=0, b=0)
)

# Build Dash app
app = dash.Dash(__name__)
app.layout = html.Div([
    html.H4("Map with Line and Hovertemplate"),
    dcc.Graph(figure=fig)
])

if __name__ == '__main__':
    app.run(debug=True)

@Hermeij

Great news @Hermeij ! Looks good on my end as well.

I'm going to close this issue, but feel free to reopen if you encounter it again.

@emilykl

@emilykl
I tried the Generic Crossfilter Recipe example from the dash docs with the latest Plotly 6.4.0 and there is still an error.

As a workaround you can fix the error by making the customdata a list
customdata=list(df.index)

Link to current open issue: plotly/plotly.py#5119. The fix might be similar to the one in pr 7608?

Navigation Menu

Repository navigation

Customdata missing in hoverdata in Ploly 6.0.0 #7375

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Customdata missing in hoverdata in Ploly 6.0.0 #7375

Description

Activity

dgibbon-plotly commented on Feb 22, 2025

Hermeij commented on Feb 23, 2025

jonasvdd commented on Mar 7, 2025

Hermeij commented on Mar 15, 2025

emilykl commented on Oct 27, 2025

Hermeij commented on Oct 29, 2025

emilykl commented on Oct 29, 2025

AnnMarieW commented on Nov 14, 2025

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions