I have a Pandas DataFrame, for example:
pd.DataFrame({
"A": [165,166,920,920,920,165,162,162,166,165,755,469],
"B": [1920,868,470,470,698,945,1670,575,701,1920,1920,470],
"C": [1670,461,461,212,212,212,212,414,453,196,254,461]
})
I would like to have a seaborn countplot of each option A, B and C with the top 2 for each column plotted side by side.
Below is an image of something similar. My problem is getting the columns side by side with the legend.
Thanks
Can you try this?
import pandas as pd
import seaborn as sns
df=pd.DataFrame({
"A": [165,166,920,920,920,165,162,162,166,165,755,469],
"B": [1920,868,470,470,698,945,1670,575,701,1920,1920,470],
"C": [1670,461,461,212,212,212,212,414,453,196,254,461]})
sns.set(style="darkgrid")
fig, ax =plt.subplots(3,1)
max_count = max([max(df[i].value_counts()) for i in df.columns])
A=sns.countplot(y=df['A'],ax=ax[0],order=df.A.value_counts().iloc[:2].index)
B=sns.countplot(y=df['B'],ax=ax[1],order=df.B.value_counts().iloc[:2].index)
C=sns.countplot(y=df['C'],ax=ax[2],order=df.C.value_counts().iloc[:2].index)
ax[0].set_xlim(0,max_count)
ax[1].set_xlim(0,max_count)
ax[2].set_xlim(0,max_count)
A.set(xticklabels=[])
B.set(xticklabels=[])
ax[0].set_title(<title>) Let me know if this works for you. :)