How to change legend labels in scatter matrix - pandas

Question

I have a scatter matrix that I want to change the labels for. On the right-hand, I want to change the blue color 1 to Say Mystery and the red color 2 to say Science. I also want to change the labels of each graph to label their counterpart [Spicy, Savory, and Sweet]. I tried using dict to relabel but then my charts came out wrong.
import plotly.express as px
fig = px.scatter_matrix(df,
dimensions=["Q12_Spicy", "Q12_Sav", "Q12_Sweet", ],color="Q11_Ans"
)
fig.show()

Answer 1

You can create a new column called Q11_Labels that maps 1 to Mystery and 2 to Science from the Q11_Ans column, and pass colors='Q11_Labels' to the px.scatter_matrix function. If you still want the legend to display the original column name, you can pass a dictionary to the labels parameter of the px.scatter_matrix function with labels={"Q11_Labels":"Q11_Ans"}
Then you can extend this dictionary to include the other column name to display name mappings as well, so that [Spicy, Savory, Sweet] are displayed instead of [Q12_Spicy, Q12_Savory, Q12_Sweet].
import numpy as np
import pandas as pd
import plotly.express as px
## recreate random data with the same columns
np.random.seed(42)
df = pd.DataFrame(
np.random.randint(0,100,size=(100, 3)),
columns=["Q12_Spicy", "Q12_Sav", "Q12_Sweet"]
)
df["Q11_Ans"] = np.random.randint(1,3,size=100)
df["Q11_Ans"] = df["Q11_Ans"].astype("category")
df = df.sort_values(by="Q11_Ans")
## remap the values of 1 and 2 to their meanings, then pass this as the color
df["Q11_Labels"] = df["Q11_Ans"].map({1: "Mystery", 2: "Science"})
## pass a dictionary to the labels parameter
fig = px.scatter_matrix(df,
dimensions=["Q12_Spicy", "Q12_Sav", "Q12_Sweet"],color="Q11_Labels",
labels = {"Q12_Spicy":"Spicy","Q12_Sav":"Savory","Q12_Sweet":"Sweet", "Q11_Labels":"Q11_Ans"}
)
fig.show()