更改 Seaborn 显示中的属性

Question

It looks like a pretty straightforward task but I cannot find the right solution. I do generate 10 collections with 100 samples in each and pack them into pandas.DataFrame assigning index numbers for each column.

How to plot all 10 lines with: same color='b' for all, set alpha=0.5, linewidth=1.0?

import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns
import pandas as pd
import random


sns.set_style("darkgrid")
sns.set_context("talk")
SEED = 202107
np.random.seed(SEED)
random.seed(SEED)


Num = 10
val = [np.random.randn() for _ in range(Num)]
my_data = [mu + np.random.randn(100) for mu in val]

df = pd.DataFrame()
for index in range(Num):
    name = str(index)
    df[name] = my_data[index][:]

sns.displot(df, kind='kde', legend=False)


plt.show()

Answer 1

The answer requires realizing that in wide-form mode , displot assigns the column dimension to a hue variable. So you control the color through the palette parameter and the easiest way to make everything one color is to pass a list of identical entries with the right length. The other properties can be controlled by matplotlib keyword arguments that get passed through to plt.plot :

g = sns.displot(
    df, kind='kde', legend=False,
    palette=["C0"] * Num, linewidth=1
)
plt.setp(g.ax.lines, alpha=.5)

It looks like there's a bug where alpha= is ignored, which you can workaround by modifying the artists after plotting (as above) or by defining the palette as [mpl.colors.to_rgba("C0", alpha=.5)] * Num .