CODEWITHSUNDEEP
pythonpandasdataframe
Giovanni Fiorillo
Giovanni Fiorillo

Reputation: 27

Select rows in data frame based all column values?

I am trying to write a python script that can take a dataframe of an arbitrary size and filter it for rows where any column matches some criteria, specifically looking for a text match. I show a 2x4 table to represent an example, but my data could have several thousand rows and anywhere from 1-15 columns.

A header Another header
Event1 Event2
Event2 OPER-8
OPER-1 Event2
Event3 Event4

I want to basically filter the dataframe to give me the rows where any entry matches the phrase "OPER*" or maybe where "Event2" is present. Most of the ways I've found searching online require you to specify a column name/number, but I want to search the entire thing. I know I can do a for loop, but are there any ways to do this more efficiently with pandas?

Upvotes: 0

Views: 34

Answers (1)

ThePyGuy
ThePyGuy

Reputation: 18416

You can apply a function on row axis of the dataframe to create True/False value for the provided condition, then take out the values based on this True/False value maksing

>>> df[df.apply(lambda x: x.str.match('^OPER-\d+|Event2$').any(), axis=1)]
  A header Another header
0   Event1         Event2
1   Event2         OPER-8
2   OPER-1         Event2

Upvotes: 2

Related Questions