将Series转换为熊猫词典列表

Question

I have a Series of the form ["A","B","B","C","A"] .

I am counting the frequency of each unique element using pandas.Series.value_counts which also returns a Series: ["A":2,"B":2,"C":1]

I want to convert this Series into a list of dictionaries [{A:2},{B:2},{C:1}] . However, neither to_dict nor to_records is giving me the desired result. What's an alternative?

Answer 1

I don't think any type-conversion method in Pandas for Series or DataFrames will produce exactly the output you're after.

You might simply have to adjust the result of to_dict() with a list comprehension:

>>> counts = pd.value_counts(series).to_dict()
>>> [{u: v} for (u, v) in counts.iteritems()]
[{'A': 2}, {'C': 1}, {'B': 2}]

Answer 2

you need counter

>>> from collections import Counter
>>> [ {x:y} for x,y in Counter(["A","B","B","C","A"]).items() ]
[{'A': 2}, {'C': 1}, {'B': 2}]