Change plotly express "color" variable with button

Question

I would like to create a plotly plot where I can change the value of the plotly express color argument via a button. I am using plotly.express.scatter for this.

For example, the initial plot shown is px.scatter(df, "sepal_length", "sepal_width", color="species"). Changing from "species" to "petal length" in a dropdown menu would update the plot so that instead color="petal_length". If it makes a difference, "species" uses the default discrete color sequence while "petal_length" uses the default continuous color scale.

The code I have so far makes the initial plot and dropdown buttons, but choosing the buttons has no effect. I don't understand how to get the plotly express color argument passed through this Plotly.update interface.

import plotly.express as px
import pandas as pd


df = pd.read_csv("https://raw.githubusercontent.com/mwaskom/seaborn-data/master/iris.csv")
fig = px.scatter(df, "sepal_length", "sepal_width", color="species")

fig.update_layout(
    updatemenus=[
        dict(
            buttons=list([
                dict(
                    args=["color", "species"],
                    label="species",
                    method="update"
                ),
                dict(
                    args=["color", "petal_length"],
                    label="petal length",
                    method="update"
                ),
            ]),
            showactive=True,
            x=0.05,
            xanchor="left",
            y=1.06,
            yanchor="top"
        ),
    ]
)

fig.update_layout(
    annotations=[
        dict(text="color", x=0.015, xref="paper", y=1.05, yref="paper",
             align="left", showarrow=False),
    ])

fig.show()

r-beginners · Accepted Answer · 2022-11-01 04:41:48Z

For color='species', three graphs are internally created for each categorical variable. And color='petal_length' consists of a single graph data. So, it is possible to handle this by setting the drop-down to show/hide the graph. The actual code reuses the data created by express.scatter. The data for each graph will be configured in a graph object; three will be shown and one will be hidden. Set the button to restyle as a function of the button.

import plotly.express as px
import plotly.graph_objects as go
import pandas as pd

df = pd.read_csv("https://raw.githubusercontent.com/mwaskom/seaborn-data/master/iris.csv")
fig1 = px.scatter(df, "sepal_length", "sepal_width", color="species")
fig2 = px.scatter(df, "sepal_length", "sepal_width", color="petal_length")

fig = go.Figure()
fig.add_trace(go.Scatter(fig1.data[0], visible=True))
fig.add_trace(go.Scatter(fig1.data[1], visible=True))
fig.add_trace(go.Scatter(fig1.data[2], visible=True))

fig.add_trace(go.Scatter(fig2.data[0], visible=False))

fig.update_layout(
    updatemenus=[
        dict(
            buttons=list([
                dict(
                    args=["visible", [True,True,True,False]],
                    label="species",
                    method="restyle"
                ),
                dict(
                    args=["visible", [False,False,False,True]],
                    label="petal length",
                    method="restyle"
                ),
            ]),
            showactive=True,
            x=0.05,
            xanchor="left",
            y=1.2,
            yanchor="top"
        ),
    ]
)

fig.update_layout(
    annotations=[
        dict(text="color", x=0.01, xref="paper", y=1.16, yref="paper",
             align="left", showarrow=False),
    ])
fig.update_layout(xaxis_title_text='sepal_length', yaxis_title_text='sepal_width', legend_title_text='species')
fig.show()

This works great, thanks! I was easily able to add a sepal_width choice too. I created a fig3 and added its trace like what you did with fig2. For the visible args in the button dicts I added new Falses at the ends for the first two and the new one is [False,False,False,False,True].

Collectives™ on Stack Overflow

Change plotly express "color" variable with button

1 Answer 1

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related