Difference Between Choropleth, Choroplethmap, and Choroplethmapbox in Plotly?

📊 Plotly Python
1

I’ve been working with plotly.graph_objects to create visualizations and have been using the Choroplethmap object. However, I couldn’t find specific documentation for Choroplethmap and I’m wondering if someone could clarify the differences between Choropleth, Choroplethmap, and Choroplethmapbox. Are there any notable distinctions, and which of these is most appropriate when working with graph_objects?

Right now, I’m using Choroplethmap for my visualizations, but it would be helpful to understand if this is the best choice or if one of the other options would be more suitable.

Any insights or references to documentation would be greatly appreciated!

Thanks in advance!

2

Hi @yumoto

To answer your second question, the difference between Choroplethmap and Choroplethmapbox is that the latter does not use MapLibre tiles, while the former does.

‘mapbox’ suffix is replaced by ‘map’ in plotly==5.24.0 see;
mapbox to maplibre
I believe that Mapbox will eventually be deprecated.

I recently changed from Choropleth to Choroplethmap for a few visualizations because I wanted to capture and use user interactions, such as clickData and relayoutData. You can overlay tiles in both but the latter allows for more interaction. The disadvantage to this is that it’s hard to get a truly minimal basemap, carto-positron-nolabels will still show arterial roads etc… My workaround was to create a custom tile using MapTiler and remove literally everything!

Here’s some code I adapted from a project that shows the main difference (both are maps of France)

Choropleth:

import plotly.graph_objects as go

# Assume `df` is a GeoDataFrame and 'data_column' contains the data to visualize
# Example: `df['population_density']` as the data column

# Step 1: Initialize a basic Plotly figure
fig = go.Figure()

# Step 2: Add the choropleth map trace
fig.add_trace(
    go.Choropleth(
        geojson=df.__geo_interface__,  # GeoJSON representation of the dataframe
        z=df['data_column'],  # Generic data to be visualized (e.g., population, density)
        locations=df.index,  # Locations mapped via the index of the dataframe
        colorscale='Blues',  
        colorbar_title='Data',  
        marker_line_width=0.5,  
        marker_line_color='darkgray',  
        hovertemplate='<b>Location:</b> %{customdata[0]}<br><b>Data:</b> %{z}<extra></extra>',  # Simple hover template
        customdata=df[['region']].values  # Extra data for the hovertemplate
    )
)

# Step 3: Set default map center (e.g., France) and zoom level
avg_lat, avg_lon = 46.603354, 1.888334  
zoom_level = 6  

# Step 4: Update layout, set map boundaries, and center it
fig.update_layout(
    geo=dict(
        scope='europe',  # Scope set to Europe
        resolution=50,  # Resolution of map boundaries
        showcoastlines=False,  
        showland=True, 
        landcolor="lightgray", 
        center=dict(lat=avg_lat, lon=avg_lon),  # Center the map on France
        projection_scale=zoom_level  # Set zoom level
    ),
    margin=dict(l=10, r=10, t=30, b=10)  # Adjust margins for less white space
)

# Step 5: Show the figure
fig.show()

Choroplethmap:

import plotly.graph_objects as go

# Assume `df` is a GeoDataFrame and 'data_column' contains the data to visualize
# Example: `df['population_density']` as the data column

# Step 1: Initialize a basic Plotly figure
fig = go.Figure()

# Step 2: Add a choropleth map trace to the figure
fig.add_trace(
    go.Choroplethmap(
        geojson=df.__geo_interface__,  # GeoJSON representation of the GeoDataFrame
        z=df['data_column'],  # The data you want to visualize, e.g., population density
        locations=df.index,  # Map locations to the index of the dataframe
        colorscale='Blues',  # Use a basic colormap
        colorbar_title='Data',  # Title for the color bar
        marker_line_width=0.5,  # Thin line for boundaries
        marker_line_color='darkgray',  # Boundary color
        hovertemplate='<b>Location:</b> %{customdata[0]}<br><b>Data:</b> %{z}<extra></extra>',  # Hover info template
        customdata=df[['region']].values  # Extra data to display in the hovertemplate
    )
)

# Step 3: Add a scatter text trace to display labels
centroids = df.geometry.centroid  # Calculate centroids for label placement
fig.add_trace(
    go.Scattermap(
        lon=centroids.x,  # Longitude of the centroids
        lat=centroids.y,  # Latitude of the centroids
        text=df['region'],  # Text labels, e.g., regions or departments
        mode='text',  # Display as text
        textposition='top center',  # Position the text
        showlegend=False  # No legend for text labels
    )
)

# Step 4: Set map layout with centering and zoom
fig.update_layout(
    map=dict(
        style="../assets/basicTileMap.json",  # Basic tile map style - the custom one created on MapTiler
        center={'lat': 46.603354, 'lon': 1.888334},  # Center the map on France
        zoom=6  # Set zoom level
    ),
    margin=dict(l=10, r=10, t=30, b=10),  # Adjust margins
    height=700  # Set figure height
)

# Step 5: Show the figure
fig.show()

If you want to see the difference live, check out Michelin Guide to France. The ‘region’ plot is Choropleth the ‘economic health’ plot is Choroplethmap.

Good luck with your project,

pineapple-bois

1 Like