How to add several new columns to a dataframe acording to some rules

user_16348663 0
Viewed 20

I have a daframe with several columns but some of them starts with test_

Below a sample with ONLY these test_ columns:

c = pd.DataFrame({'test_pierce':[10,30,40,50],'test_failure':[30,10,20,10] })

What I need to do:

For every column in my dataframe that starts with test_ I want to create another column just after to classify it's value like this:

if test_ > 30.0:
   Y
else:
   N

To get this output:

d = pd.DataFrame({'test_pierce':[10,30,40,50],'class_test_pierce':['N','N','Y','Y'],'test_failure':[30,10,20,10], 'class_test_failure':['N','N','N','N'] })

What I did:

I have the columns I need to classify:

cols = [c for c in c.columns if c.startswith('test_')]

I couldn't proceed from here tho

jupyter-notebook pandas
2 Answers

Code with the suggested order: The code is a little ugly because you asked to be the columns after its test_ column. Otherwise, the code is simpler than that.

cols = [(i,c) for i,c in enumerate(c.columns) if c.startswith('test_')]

count = 1 
for index,col in cols:
    value = np.where(c[col] > 30.0,'Y','N')
    c.insert(index+count, 'class_'+col, value)
    count +=1

enter image description here

Code without the suggested order:

cols = [c for c in c.columns if c.startswith('test_')]

for col in cols:
    c[f'class_{col}'] = np.where(c[col] > 30.0,'Y','N')
user_167331010

A format that may help you get started is:

cols = [c for c in c.columns if c.startswith('test_')]

for col in cols:
    df[f'class_{col}'] = df.apply(lambda x: 'Y' if x[col] > 30.0 else 'N', axis=1)

Output:

   test_pierce  test_failure class_test_pierce class_test_failure
0           10            30                 N                  N
1           30            10                 N                  N
2           40            20                 Y                  N
3           50            10                 Y                  N
user_118659560
Related