Conditions in for loop

Using pandas data frame in python:

I am trying to get the following data from a data frame and cant fix my loop to get the proper result.

A small sample of the data set:

BERTH     FROM_BERTH       BI
29          H2             0
29          09             0
29          J5             0
C5          NaN            1
J4          NaN            1
J2          NaN            1

Here is part of the dataset ( the relevant columns i need to use) and here is what i want the output to look like :

29, H2, 09, J5, C5, J4, J2

that is: if BI is 0 i want the value from BERTH and then the values of FROM_BERTH till BI becomes 1, and then goes back to giving me the BERTH ( untill BI becomes 1 again, so on ans so forth ).

Here are the following codes i have tried and their results :

test_berth2 = []

for i in range(0,6):
    if df3_test.loc[i,'BI'] == 0 & df3_test.iloc[i,21] != df3.iloc[i-1,21]:
        test_berth2.append(df3_test.loc[i,'BERTH'])
        test_berth2.append(df3_test.loc[i,'FROM_BERTH'])
    elif df3_test.loc[i,'BI'] == 0 & df3_test.iloc[i,21] == df3.iloc[i+1,21]:         
        test_berth2.append(df3_test.loc[i,'FROM_BERTH'])

    else :
        test_berth2.append(df3_test.loc[i,'BERTH'])

test_berth2

result :

['29', 'H2', '29', '09', '29', 'J5', 'C5', 'J4', 'J2']

Loop 2 :

for i in range(0,6):
    if df3_test.iloc[i,21] == 0:
        print (df3_test.loc[i,'BERTH'])
        while df3_test.iloc[i,21] == 0:
            print (df3_test.loc[i,'FROM_BERTH'])
            i = i+1
    else:
        print (df3_test.loc[i,'BERTH'],'1')

result :

29, H2, 09, J5, 29, 09, J5, 29, J5, C5 1, J4 1, J2 1

PS :the iloc to column 21 is 'BI' btw