lunny/keras
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Make pre-padding the default in sequence tensors

Browse Source
This commit is contained in:
fchollet 2015-06-18 11:26:11 -07:00
parent 4830b4be27
commit 6329378ca3
1 changed files with 7 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
keras/preprocessing/sequence.py
View File

@ -4,13 +4,15 @@ import numpy as np
import random import random
from six.moves import range from six.moves import range
def pad_sequences(sequences, maxlen=None, dtype='int32'): def pad_sequences(sequences, maxlen=None, dtype='int32', padding='pre'):
""" """
Pad each sequence to the same length: Pad each sequence to the same length:
the length of the longuest sequence. the length of the longuest sequence.
If maxlen is provided, any sequence longer If maxlen is provided, any sequence longer
than maxlen is truncated to maxlen. than maxlen is truncated to maxlen.
Support post-padding and pre-padding (default).
""" """
lengths = [len(s) for s in sequences] lengths = [len(s) for s in sequences]
@ -20,7 +22,10 @@ def pad_sequences(sequences, maxlen=None, dtype='int32'):
x = np.zeros((nb_samples, maxlen)).astype(dtype) x = np.zeros((nb_samples, maxlen)).astype(dtype)
for idx, s in enumerate(sequences): for idx, s in enumerate(sequences):
x[idx, :lengths[idx]] = s[:maxlen] if padding == 'post':
x[idx, :lengths[idx]] = s[:maxlen]
else:
x[idx, -min(maxlen, lengths[idx]):] = s[:maxlen]
return x return x