Count the occurrence of one column based on another

Question

I have a dataframe like this

System	package
mac	abc
mac	bcd
windows	bcd
mac	abc

I want a code in python to get a dataframe like this

System	count of distinct package
mac	2
windows	1

Answer 1

Create dataframe:

pd.DataFrame(data={"a":["mac","mac","windows","mac",],"b": ["abc", "bcd", "bcd", "abc"]})

group it:

t.groupby('a').b.nunique()

output:

    a
mac        2
windows    1
Name: b, dtype: int64