CODEWITHSUNDEEP
pythondataframe
Milrayitas
Milrayitas

Reputation: 35

Get last row with specific value in column, Python

I have

df = pd.DataFrame('data', car, 'ok')

df:

 data               car         ok                                                       
2020-03-25           A            
2020-04-01           A           x
2020-04-15           A              
2020-05-08           A           x
2020-06-25           A           x
2020-06-27           A
2020-07-15

I want to select last (old in this case) row being column 'ok' with "x"

I want to obtain

2020-04-01   A     x

Thanks!

Upvotes: 1

Views: 3410

Answers (2)

sparc_spread
sparc_spread

Reputation: 10833

The head() method on DataFrame can give you the first n rows of a DataFrame. Indexing of a DataFrame will allow you to select rows meeting your filter criteria - to narrow to a subset DataFrame with such rows. Together, you could use them to do:

r = df.loc[df.ok == 'x', :].head(1)

What you are doing here is narrowing to a subset DataFrame where ok is 'x' (the df.loc[df.ok == 'x', :] part), then taking the first row of it (the .head(1) part). This of course assumes the DataFrame is sorted by date as it is above.

Indexing is a huge and fundamental topic (think of it as the SQL WHERE of pandas) so you should spend time gaining a deep knowledge of it. Here is a great tutorial. This one and this one are also good.

Upvotes: 4

Alexander Golys
Alexander Golys

Reputation: 703

This will work also when your data is not sorted:

df[df.ok == 'x'][df.data == df.data.min()]

Upvotes: 1

Related Questions