How to describe columns as categorical values?

Question

I have a pandas dataframe that contains a mix of categorical and numeric columns. By default, df.describe() returns only a summary of the numerical data (describing those columns with count, mean, std, min, quantiles, max)

when iterating through all the columns in the df and describing them individually as [df[c].describe() for c in df.columns] the description is returned based off of specific column dtype; i.e. numerical summary for int and float and categoric summary for object

Does any one know of a succinct way of describing all columns as categorical with count, unique, top, freq?

I think I answered myself while clarifying the question. just tried `df.astype('object').describe()` and it works!! — Zahra, Feb 26 '18 at 21:49

score 15 · Answer 1 · answered Dec 04 '18 at 12:40

15

a slightly shorter version of the answer:

df.describe(include = 'object')

answered Dec 04 '18 at 12:40

IndPythCoder

693
6
10

score 13 · Accepted Answer · answered Feb 26 '18 at 21:54

13

following converts all columns to object type then describes them:

df.astype('object').describe()

for cleaner view try:

df.astype('object').describe().transpose()

answered Feb 26 '18 at 21:54

Zahra

6,798
9
51
76

2

Thank you so much! By the way, for shorter code, try `T` instead of `transpose()` – Arthur Khazbs Jun 29 '22 at 16:01

score 0 · Answer 3 · answered Dec 10 '22 at 07:25

0

Additionally, if YOU want to see descriptions of all columns together(including categorical). you can use:

df.describe(include='all')

answered Dec 10 '22 at 07:25

Saurav

163
2
4

How to describe columns as categorical values?

3 Answers3