Search code examples
pythonseabornhistogram

Seaborn stacked histogram with data from multiple columns

How can I can plot multiple stacked histograms using Seaborn? I tried the following code, but it threw a dimensions error: ValueError: Length of list vectors must match length of data...

df = pd.DataFrame({'id': [1,2,3,4,5,6,7,8,9,10],
                   'val1': ['a','b',np.nan,np.nan,'a','a',np.nan,np.nan,np.nan,'b'],
                   'val2': [7,0.2,5,8,np.nan,1,0,np.nan,1,1],
                   'cat': ['yes','no','no','no','yes','yes','yes','yes','no','yes'],
                  })
display(df)

sns.histplot(data=df, y=['val1', 'val2'], hue='cat', multiple='stack')

enter image description here

Desired Plot:
enter image description here
val1 "no" freq = 1 and "yes" = 4
val2 "no" freq = 4 and "yes" = 4

Solution

  • I don't think you'll be able to do this directly from your current data frame. You need to get a dataframe that has val1/val2 in one column and yes/no in another.

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

df = pd.DataFrame({'id': [1,2,3,4,5,6,7,8,9,10],
                   'val1': ['a','b',np.nan,np.nan,'a','a',np.nan,np.nan,np.nan,'b'],
                   'val2': [7,0.2,5,8,np.nan,1,0,np.nan,1,1],
                   'cat': ['yes','no','no','no','yes','yes','yes','yes','no','yes'],
                  })

val1 = df[['cat', 'val1']].dropna().drop(columns='val1')
val1['val'] = 'val1'

val2 = df[['cat', 'val2']].dropna().drop(columns='val2')
val2['val'] = 'val2'

plot_df = val1.append(val2).sort_values(by='cat')

sns.histplot(data=plot_df,x='val', stat='count', hue='cat', multiple='stack')

plt.show()