Fill new column with True/False for each unique customer based on condition-CodePudding

I need to create new column with True / False values. If condition is met based on other column, than True should be in every row associate with acc_no.

Original df:

trans_id acc_no  trans_amount     
   1       123        15     
   2       456        10
   3       123        50     
   4       456        60     
   5       789        35
   6       123        20          
   7       789        18          
   8       456        73
   9       789        37

output should be True for 123 acc_no only:

trans_id acc_no  trans_amount ANY_50$_TRANSACTION ?    
   1       123        15              True
   2       456        10              False           
   3       123        50              True     
   4       456        60              False     
   5       789        35              False
   6       123        20              True          
   7       789        18              False         
   8       456        73              False
   9       789        37              False

My code:

any_50$_transaction = []

for i in df["acc_no"].unique():
    
    for ind in df[df["acc_no"] == i].index:
    
        any_50$_transaction.append(np.where(df['trans_amount'][ind] == 50, True, False).any())       
        
df["ANY_50$_TRANSACTION ?"] = any_50$_transaction 

df

So far my output is:

trans_id acc_no  trans_amount ANY_50$_TRANSACTION ?    
   1       123        15              False
   2       456        10              False           
   3       123        50              True     
   4       456        60              False     
   5       789        35              False
   6       123        20              False
   7       789        18              False         
   8       456        73              False
   9       789        37              False

CodePudding user response：

Use Series.isin with filtered acc_no by condition:

df['ANY_50$_TRANSACTION'] = df['acc_no'].isin(df.loc[df['trans_amount'].eq(50), 'acc_no'])

Or GroupBy.transform with GroupBy.any:

df['ANY_50$_TRANSACTION'] = df['trans_amount'].eq(50).groupby(df['acc_no']).transform('any')

print (df)
   trans_id  acc_no  trans_amount  ANY_50$_TRANSACTION
0         1     123            15                 True
1         2     456            10                False
2         3     123            50                 True
3         4     456            60                False
4         5     789            35                False
5         6     123            20                 True
6         7     789            18                False
7         8     456            73                False
8         9     789            37                False