using apply function write multiple append statement

January 8, 2022

Is it possible to write below code using apply function or any other, because iterrows seems not feasible here.

For creating Dataframe:

df = [[['55792', 'TRIM'], 'A', ['55792', 'TRIM']], [['ZQFC', 'DYWH'], 'A', ['MEIL', '1724']]]
df = pd.DataFrame(df, columns=['seg','cod','seg2'])
df 

seg            cod  seg2
[55792, TRIM]   A   [55792, TRIM]
[ZQFC, DYWH]    A   [MEIL, 1724]


#output
seg            cod  seg2
[55792, TRIM]   A   [55792, TRIM]
[ZQFC, DYWH]    A   [MEIL, 1724]
[MEIL, 1724]    A   [MEIL, 1724]

So, I am expanding the rows if the seg column and seg2 column doesn’t matches. It there any better way to do that using apply.

Code I have used:

df1 = pd.DataFrame(columns=df.columns)
for index, row in df.iterrows():
    #if seg match seg 2
    if(row['seg']==row['seg2'])==True:
        #append same row
        df1 = df1.append(row)
    else:
        #if not match, then we can create two seg, and append 2 row
        #append same row first
        df1 = df1.append(row)
        #changing existing segment to new segment, and append
        row['seg'] = row['seg2']
        df1 = df1.append(row)

Thanks a lot.

>Solution :

You don’t need to use apply you can just use your conditionals as boolean masks and do your operations that way.

mask = df["seg"] == df["seg2"]

true_rows = df.loc[mask]
false_rows = df.loc[~mask]
changed_rows = false_rows.assign(seg=false_rows.seg2)

df1 = pd.concat([true_rows, false_rows, changed_rows], ignore_index=True)

print(df1)
             seg cod           seg2
0  [55792, TRIM]   A  [55792, TRIM]
1   [ZQFC, DYWH]   A   [MEIL, 1724]
2   [MEIL, 1724]   A   [MEIL, 1724]