Getting a list of the range of 2 pandas columns

Question

I have the following DataFrame (reformatted a bit):

    f_name  l_name      n   f_bought  l_bought
0   Abraham Livingston  24  1164      1187
1   John    Brown       4   1188      1191
2   Samuel  Barret      16  1192      1207
3   Nathan  Blodget     4   1208      1212
4   Bobby   Abraham     1   1212      1212

I want to create a column, bought , that is a list range(df[f_bought], df[l_bought]) .

I've tried:

def getRange(l1,l2):
    r = list(range(l1, l2))
df.apply(lambda index: getRange(df['f_bond'], df['l_bond']),axis=1)

but it results in a TypeError:

"cannot convert the series to <type 'int'>", u'occurred at index 0'

I've tried a df.info(), and both columns are type int64.

I'm wondering if I should use something like df.loc[] or similar? Or something else entirely?

Answer 1

You should be able to do this using apply which is for applying a function to every row or every column of a data frame.

def bought_range(row):
    return range(row.f_bought, row.l_bought)

df['bought_range'] = df.apply(bought_range, axis=1)

Which results in:

f_name      l_name   n  f_bought  l_bought  \
0  Abraham  Livingston  24      1164      1187   
1     John       Brown   4      1188      1191   
2   Samuel      Barret  16      1192      1207   
3   Nathan     Blodget   4      1208      1212   
4    Bobby     Abraham   1      1212      1212   

                                        bought_range  
0  [1164, 1165, 1166, 1167, 1168, 1169, 1170, 117...  
1                                 [1188, 1189, 1190]  
2  [1192, 1193, 1194, 1195, 1196, 1197, 1198, 119...  
3                           [1208, 1209, 1210, 1211]  
4                                                 []

One word of warning is that Python's range doesn't include the upper limit:

In [1]: range(3, 6)
Out[1]: [3, 4, 5]

It's not hard to deal with ( return range(row.f_bought, row.l_bought + 1) ) but does need taking into account.