Reputation: 1973
Is there a way to plot the CDF + cumulative histogram of a Pandas Series in Python using Seaborn only? I have the following:
import numpy as np
import pandas as pd
import seaborn as sns
s = pd.Series(np.random.normal(size=1000))
I know I can plot the cumulative histogram with s.hist(cumulative=True, normed=1)
, and I know I can then plot the CDF using sns.kdeplot(s, cumulative=True)
, but I want something that can do both in Seaborn, just like when plotting a distribution with sns.distplot(s)
, which gives both the kde fit and the histogram. Is there a way?
Upvotes: 60
Views: 69002
Reputation: 49032
import numpy as np
import seaborn as sns
x = np.random.randn(200)
stat = "count" # or proportion
sns.histplot(x, stat=stat, cumulative=True, alpha=.4)
sns.ecdfplot(x, stat=stat)
Upvotes: 107
Reputation: 1982
You can get almost the same plot using matplotlib by using cumulative=True
and density=True
.
plt.hist(x,cumulative=True, density=True, bins=30)
Upvotes: 11