pandas series: why series[-1] = error and series[0] = result-CodePudding

Given a series like pd.Series([1,2,3,4]) To get the first value I can do:

first = series[0]

But to get the last value like last = series[-1] produces an error. I need to do last = series.iloc[-1] I don't understand why. (Why do I don't need first = series.iloc[0] or can do last = series[-1]?

Thank you.

CodePudding user response：

To answer your question, let's do a little experiement.

s = pd.Series([1,2,3,4],index=[9,8,7,6]) 
s.index

Output:

Int64Index([9, 8, 7, 6], dtype='int64')

If you have this series, even S[0] will give an error. But you can do S[9], S[8], etc..

s[9]
1
s[8]
2
s[0]
KeyError: 0
s[-1]
KeyError: -1

When you are using s[0], you are using the index value which does not exist. Same for s[-1] in this case.

When you want to use -1 to refer to the last item, you can still do it but like this:

s = pd.Series([1,2,3,4])
s[-1:]

Output:

3    4
dtype: int64

You are telling the series, you want the last item till the end, not specifing the actual index value. In this case, 3 is the index, 4 is the value.

And if you want to refer to the 2nd last item, like this:

s = pd.Series([1,2,3,4])
s[-2:-1]

Output:

2    3
dtype: int64