Python Pandas, check the condition of certain columns in a row?

Question

I have a dataframe:

      var1  var2  var3  var4
Id#                         
1001     Y     Y     Y     Y
1002     N     N     N     N
1003     N     N     Y     N
1003     Y     Y     Y     N

I want to create a new column called Small, where if any var=Y then Small is equal to N

      var1  var2  var3  var4  Small
Id#                         
1001     Y     Y     Y     Y      N       
1002     N     N     N     N      Y 
1003     N     N     Y     N      N
1003     Y     Y     Y     N      N

My tried solution: I have created a function called is_small that flips to 'N' anytime there a column in a row is 'Y'

def is_small(row, *cols):
    _small = 'Y'
    for col in cols:
        if col == 'Y':
            _small = 'N'
    return _small

and apply it to my dataset:

all_data['Small'] = all_data.apply(lambda row: is_small(row,
                                                        'var1',
                                                        'var2',
                                                        'var3',
                                                        'var4'),
                                   axis=1)

However Small just all return as 'Y' and I'm not sure why.

gzc · Accepted Answer · 2017-05-05 15:40:46Z

1

You are almost there. but every time you pass literal 'var1', 'var2'... into is_small, that's why it always return 'Y'. You should pass row['var1'], row['var2']....

all_data['Small'] = all_data.apply(lambda row: is_small(row,
    row['var1'],
    row['var2'],
    row['var3'],
    row['var4']),
axis=1)

answered May 5, 2017 at 15:40

gzc

8,6998 gold badges44 silver badges63 bronze badges

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

1 Comment

spitfiredd Over a year ago

So what I was doing was passing my function the list of *args ['var1', 'var2', 'var3', 'var4']?

akuiper · Accepted Answer · 2017-05-05 14:26:30Z

1

You can use numpy.where (vectorized if/else):

df['small'] = pd.np.where(df.eq('Y').any(1), 'N', 'Y')

df
#    var1 var2  var3  var4  small
#Id                 
#1001   Y    Y     Y    Y   N
#1002   N    N     N    N   Y
#1003   N    N     Y    N   N
#1003   Y    Y     Y    N   N

answered May 5, 2017 at 14:26

akuiper

216k33 gold badges362 silver badges379 bronze badges

2 Comments

spitfiredd Over a year ago

I have other columns besides var1, var2, var3, var4., that may have Y/N responses.

akuiper Over a year ago

You can select these four columns manually before constructing the boolean condition if that is the case. pd.np.where(df[['var1', 'var2', 'var3', 'var4']].eq('Y').any(1), 'N', 'Y')

Collectives™ on Stack Overflow

Python Pandas, check the condition of certain columns in a row?

2 Answers 2

1 Comment

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related