Delete rows from python panda dataframe

Question

My dataframe has columns like ticket, host, drive model, Chassis, Rack, etc.

I want all the rows with value in the Chassis column equal to '1025C-M3B', '1026T-M3FB', '2026TT-DLRF' or 'SYS-2027TR-D70RF+'. I want to delete the rest.

I tried

data2 = data1[data1.Chassis == '1025C-M3B' 
              or data1.Chassis == '1026T-M3FB' 
              or data1.Chassis == '2026TT-DLRF' 
              or data1.Chassis == 'SYS-2027TR-D70RF+']

Got

ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

Then tried

data2 = data1[data1.Chassis.all() == '1025C-M3B' 
              or data1.Chassis.all() == '1026T-M3FB' 
              or data1.Chassis.all() == '2026TT-DLRF' 
              or data1.Chassis.all() == 'SYS-2027TR-D70RF+']

Got

KeyError: u'no item named False'

Can anyone please tell me how to do this?

use bitwise or (`|`) instead of logical `or` – shx2 Jun 01 '15 at 19:52 — shx2, Jun 01 '15 at 19:52

score 2 · Answer 1 · edited May 23 '17 at 12:24

2

Use bitwise or (|) instead of logical or.

data2 = data1[(data1.Chassis == '1025C-M3B') | (data1.Chassis == '1026T-M3FB') | (data1.Chassis == '2026TT-DLRF') | (data1.Chassis == 'SYS-2027TR-D70RF+')]

You can find plenty of reading material about the use of bitwise vs logical operations in numpy/pandas. Here is one.

edited May 23 '17 at 12:24

Community

1
1

answered Jun 01 '15 at 19:54

shx2

61,779
13
130
153

unutbu · Accepted Answer · 2015-06-01T19:59:11.513

1

You could use isin:

data1[data1['Chassis'].isin(['1025C-M3B', '1026T-M3FB', '2026TT-DLRF','SYS-2027TR-D70RF+'])]

edited Jun 01 '15 at 19:59

answered Jun 01 '15 at 19:56

unutbu

842,883
184
1,785
1,677

Delete rows from python panda dataframe

2 Answers2