How to combine two relplots in seaborn python?

问题

I would like to plot two data columns of a dataframe in a single plot using sns.relplot. The dataframe looks like this:

index   x-axis  col1    col2    group   group2
0   0   27  26  A   C
1   1   45  27  B   D
2   2   48  22  A   C
3   3   35  24  B   D
4   4   49  38  A   C
5   5   46  29  B   D
6   6   29  37  A   C
7   7   38  41  B   D
8   8   24  46  A   C
9   9   46  38  B   D
10  10  37  23  A   C

Here, I want to plot col1 and col2 together against x-axis data. 'group' is the value of 'hue', and 'group2' for 'col' in the relplot.

I am able to plot the two columns separately using two individual relplots.

Plot of col1

Plot of col2

I would like to combine the two plots such that there is one single plot containing col1 and col2.

回答1:

You can melt your DataFrame and use the resulting variable as a style grouping:

from io import StringIO
import numpy as np
import pandas as pd
import seaborn as sns

data = """index   x-axis  col1    col2    group   group2
0   0   27  26  A   C
1   1   45  27  B   D
2   2   48  22  A   C
3   3   35  24  B   D
4   4   49  38  A   C
5   5   46  29  B   D
6   6   29  37  A   C
7   7   38  41  B   D
8   8   24  46  A   C
9   9   46  38  B   D
10  10  37  23  A   C"""

df = pd.read_csv(StringIO(data), index_col=[0], sep=" ", skipinitialspace=True)

sns.relplot(
    data=df.melt(id_vars=["x-axis", "group", "group2"], value_vars=["col1", "col2"]),
    x="x-axis", y="value", style="variable", hue="group", col="group2", kind="line")

Output:

来源：https://stackoverflow.com/questions/58034473/how-to-combine-two-relplots-in-seaborn-python