Skimpy a better way to analyze your df than `.describe()`

09.11.2022 — Python, Data — 1 min read

Another Python and Data post today for you, by me, with the help of what I read on Twitter ... and used. Now it's time to share it with you too!

What is skimpy?

skimpy is a light weight tool that provides summary statistics about variables in data frames within the console. -- Philip Vollet

I tested it just after seeing the tweet. And I could say it's great and I like it! Yes, it's no news that I like data and this library from Python could be used for SEO purposes.

How to install skimpy?

You can:

go to https://github.com/aeturrell/skimpy
do a pip install
load it thanks to from skimpy import skim
test it on your dataframe skim(df)

These steps are:

1!pip install
2from skimpy import skim
3skim(df)

Where df will be your dataframe with you own data. If you want to test it out of the box, there is a test data df that you can create with df = generate_test_data()

What could be the output of skimpy

This is an example I took from one of my data frame from GSC data -- as you will see with clicks and impressions from the screen capture below.

What do we learn with skimpy? More information on your dataframe and I found it more complete than .describe() from pandas.

Coming from a ... tweet of mine

This is coming from on of my tweet: https://twitter.com/ArthurCa/status/1446481101539717121