Findall across multiple dataframe columns

Question

data = {'Cat':  ['A Phaser','A','B Phaser','B','B','B'],
        'L1': ['Phase','xyzss','xyzss','Phase','xyzss','xyzss'],
        'L2': ['xyzss','Stage','xyzss','xyzss','Phase2','xyzss'],
        'L3': ['xyzss','xyzss','xyzss','xyzss','xyzss','Step'],
        }

df = pd.DataFrame (data, columns = ['Cat','L1','L2','L3'])

def funt(s):
    if re.findall(r'Phase', s, re.IGNORECASE):
        return 'Phase'
    elif re.findall(r'Stag', s, re.IGNORECASE): 
        return 'Stage'
    elif re.findall(r'Step', s, re.IGNORECASE): 
        return 'Step'
    
df[['L1','L2','L3']].apply(lambda row: '_'.join(row.values.astype(str)), axis=1).apply(lambda x : funt(x))

Output:

0    Phase
1    Stage
2     None
3    Phase
4    Phase
5     Step
dtype: object

I am wondering if there is another way of approaching this like a way of applying findall across columns without joining columns together? Thanks in advance!

Answer 1

Filter required rows. Using replace, null the xyzss. Stack and reset index and you have your outcome as a pd. Series.

Option 1: If xyzss does not vary: df['filter']=df.iloc[:,1:4].replace({'xyzss':np.nan}).stack().reset_index(drop=True)

Option 1: If xyzss varies:

df.join(pd.Series(df.mask(~df.isin(pat), np.nan).stack().reset_index(level=1, drop=True),name='filter'))