Conditional select over multiple columns - Python

Question

I have the following table loaded as a dataframe in Python using pandas

+--------+-------+------+
| Number | Col1  | Col2 |
+--------+-------+------+
| ABC    | TRUE  | SFG  |
| BCD    | TRUE  |      |
| CDE    | FALSE | SFG  |
| DEF    | FALSE |      |
| FEG    | TRUE  | JJI  |
+--------+-------+------+

Number, Col2 - String; Col1 - Boolean

I want to select rows from this df using the following logic

IF Col1 = TRUE and Col2 is not null Select Number + "," + Col2
ELSE IF Col1 = TRUE and Col2 is null Select Number
ELSE IF Col2 is not null and Col1 = FALSE Select Col2

In the above case, the output should be a list with the following values

["ABC", "SFG", "BCD", "FEG", "JJI"] //Removing the repetition too ("SFG")

How do I go about implementing this logic in Python using Pandas?

hmmm, i got ['ABC,SFG', 'BCD', 'SFG', nan, 'FEG,JJI'] following your logic — MaxU - stand with Ukraine
– MaxU - stand with Ukraine, Commented Oct 20, 2016 at 21:07
you'll probably want to use nested numpy.where statements. — Paul H
– Paul H, Commented Oct 20, 2016 at 21:18

piRSquared · Accepted Answer · 2016-10-21 05:38:48Z

2

use where + stack + tolist

pd.concat([df.Number.where(df.Col1, np.nan), df.Col2], axis=1).stack().tolist()

['ABC', 'SFG', 'BCD', 'SFG', 'FEG', 'JJI']

to get a unique list

pd.concat([df.Number[df.Col1], df.Col2.dropna()]).unique().tolist()

['ABC', 'BCD', 'FEG', 'SFG', 'JJI']

answered Oct 21, 2016 at 5:38

piRSquared

296k68 gold badges509 silver badges654 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Uğur Güney · Accepted Answer · 2016-10-20 21:46:29Z

2

Here is an implementation of your query in multiple steps:

import pandas as pd
df = pd.DataFrame(data={'Number': ['ABC', 'BCD', 'CDE', 'DEF', 'FEG'],
                        'Col1': [True, True, False, False, True],
                        'Col2': ['SFG', None, 'SFG', None, 'JJI']})
cond1 = df.Col1 & ~df.Col2.isnull()
cond2 = df.Col1 & df.Col2.isnull()
cond3 = ~df.Col1 & ~df.Col2.isnull()
selects = [df[cond1].Number + ',' + df[cond1].Col2, 
           df[cond2].Number, 
           df[cond3].Col2]
result = pd.concat(selects).sort_index()

result is (the same as @MaxU predicted)

0    ABC,SFG
1        BCD
2        SFG
4    FEG,JJI
dtype: object

answered Oct 20, 2016 at 21:46

Uğur Güney

1748 bronze badges

Collectives™ on Stack Overflow

Conditional select over multiple columns - Python

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related