Pandas If else empty

Question

Looking for an solution for if else statement in Pandas. Example:

col1  col2
1     NAN
2     NAN
3     NAN
NaN   01-2019
2     NAN

My new column needs to be col3;

When col1 = 1 or higher, add "text a"
When col1 = empty and col2 = an value, Take value of col2
Else; set value "text b"

I have now only; when col1 is higher then 1, set text a, otherwise text b.

df['col3'] = np.where(df['col1']>=1, 'text a', 'text b')

But missing the part where to check if col1 is empty and col2 has an value. To put that value inside col3.

How can i achieve that?

Thanks!

-- EDIT --

Asked also under answer for when col1 = 0 and col2 has an value, to set col3 to the value of col2.

so also:

col1  col2
0     01-2019

jezrael · Accepted Answer · 2018-10-18 06:20:07Z

3

Use numpy.select with test missing and not missing values by Series.isna and Series.notna:

print (df)
   col1     col2
0   0.0      NaN <-added row for test all coditions failed
1   1.0      NaN
2   2.0      NaN
3   3.0      NaN
4   NaN  01-2019
5   2.0      NaN

m1 = df['col1'] > =1
m2 = df['col1'].isna() & (df['col2'].notna())
#oldier pandas versions
#m2 = df['col1'].isnull() & (df['col2'].notnull())
df['col3'] = np.select([m1, m2], ['text a', df['col2']], 'text b')
print (df)
   col1     col2     col3
0   0.0      NaN   text b
1   1.0      NaN   text a
2   2.0      NaN   text a
3   3.0      NaN   text a
4   NaN  01-2019  01-2019
5   2.0      NaN   text a

Another solution with double np.where:

df['col3'] = np.where(m1, 'text a',
             np.where(m2, df['col2'], 'text b'))

EDIT:

Condition is changed:

m2 = (df['col1'] == 0) & (df['col2'].notna())

edited Oct 18, 2018 at 6:20

answered Oct 18, 2018 at 5:40

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges

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

2 Comments

HRR1337 Over a year ago

Awesome! Thanks. Can you change the answer to have also in mind when col1=0, and col2 has value, that it will take the value in col3?

HRR1337 Over a year ago

Hi there. The supplier has modified there data, and now in the stock there can also be negative values. for example in col1, -10. How can we achieve an solution that -10 also gets the col2 date? <=0 doesn't work btw.

Praveen · Accepted Answer · 2018-10-18 06:05:54Z

1

An another way to do it apart from @jexrael awesome answer is to use apply

In [27]: def condition(r):
    ...:     if r['col1'] >= 1: return "text a"
    ...:     if pd.isnull(r['col1']) and pd.notnull(r['col2']): return r['col2']
    ...:     return "text b"
    ...:

In [28]: df['col3'] = df.apply(condition, axis=1)

In [29]: df
Out[29]:
   col1     col2     col3
0   1.0      NaN   text a
1   2.0      NaN   text a
2   3.0      NaN   text a
3   NaN  01-2019  01-2019
4   2.0      NaN   text a

answered Oct 18, 2018 at 6:05

Praveen

9,4634 gold badges38 silver badges51 bronze badges

Collectives™ on Stack Overflow

Pandas If else empty

2 Answers 2

2 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related