I am trying to create an Upset Plot using following code, but my output is not as expected. See Image1: Output which I am getting vs Image2: Output expected. Where the total count of overlap is displayed at the top of each bar, Bar size is proportionate with the count. it's in descending order. Please help me with this. Thank you.

The code is 

import pandas as pd

pip install upsetplot

from upsetplot import UpSet

import matplotlib.pyplot as plt

df_spark=spark.table("hive_metastore.sndbx_cx.q3_data")

df_spark.show(5)

df=df_spark.toPandas()

df=df.drop(columns=['member_id'])

df_counts=df_multi.value_counts()

df=df.astype(bool)

from upsetplot import UpSet

df_multiindex=df.groupby(list(df.columns)).size()

fig=plt.figure(figsize=(15,10),dpi=150)

upset=UpSet(df_multiindex,subset_size='count',show_counts=True,min_subset_size=1,max_subset_size=15)
upset.plot()

plt.show()