Pandas NaN Error in creating a series with dict

import pandas as pd
pd.Series(data={"a":[1,2,3,3,5]}, index=[1,2,3,4,5])

Problem description

If I use the code above, the output is like this.

1   NaN
2   NaN
3   NaN
4   NaN
5   NaN
dtype: float64

According to the document,it says data could be dict, but actually, if I use dict, the data in Series object will be NaN.

I don't know why ? Is the way I proposed above to create a series is illegal?

Comment From: TomAugspurger

See the second section here about from a dictionary. The keys are interpreted as the index, and the values are the values of the Series, so you get a Series with a single row, A, and then reindex it. Essentially:

In [3]: pd.Series([[1, 2, 3, 3, 5]], index=['A'])
Out[3]:
A    [1, 2, 3, 3, 5]
dtype: object

In [4]: pd.Series([[1, 2, 3, 3, 5]], index=['A']).reindex([1, 2, 3, 4, 5])
Out[4]:
1    NaN
2    NaN
3    NaN
4    NaN
5    NaN
dtype: object

The docstring could probably use some examples, but we have other issues for that so I'm going to close this one.