filtering pandas dataframe with array of unique values

Question

I have a pandas DF with two columns, one columns consists of unique filenames and the other consists of that filenames' label. I also have a numpy array of the filenames that i can use for training my model. I need to extract the labels from the DF that match the filenames i can use. I tried this:

x = []

for i in nparray:
  for j in DF['filenames']:
    if DF['filenames'][j] == nparray[i]:
      x.append(DF['label'][j])

if I do this i get a key error with the name of the first filename of the DF

Please add input data and expected output – Dani Mesejo Dec 04 '19 at 11:50 — Dani Mesejo, Dec 04 '19 at 11:50

Prince Francis · Accepted Answer · 2019-12-04T11:58:51.803

1

try like these

DF[DF['filenames'].isin(nparray)]

To get only label column

DF[DF['filenames'].isin(nparray)]['label']

edited Dec 04 '19 at 11:58

answered Dec 04 '19 at 11:54

Prince Francis

2,995
1
14
22

score 1 · Answer 2 · answered Dec 04 '19 at 11:59

1

list(DF[DF['filenames'].isin(nparray)]['label'])

Will give you a list with labels in DF where it's value in filenames is in nparray

answered Dec 04 '19 at 11:59

Aryerez

3,417
2
9
17

Thanks, this also works – Stijn Roovers Dec 04 '19 at 16:29

filtering pandas dataframe with array of unique values

2 Answers2