Package pylearn :: Package datasets :: Module smallNorb
[hide private]

Module smallNorb

source code

Classes [hide private]
Paths
File-related operations on smallNorb
Functions [hide private]
 
smallnorb_iid(ntrain=29160, nvalid=9720, ntest=9720, dtype='float64', normalize_pixels=True)
Variation of the smallNorb task in which we randomly shuffle all the object instances together before dividing into train/valid/test.
source code
 
smallnorb_azSplit() source code

Imports: os, numpy, read, data_root


Function Details [hide private]

smallnorb_iid(ntrain=29160, nvalid=9720, ntest=9720, dtype='float64', normalize_pixels=True)

source code 

Variation of the smallNorb task in which we randomly shuffle all the object instances together before dividing into train/valid/test.

The default train/valid/test sizes correspond to 60/20/20 split of the entire dataset.

:returns: 5, (train_x, train_labels), (valid_x, valid_labels), (test_x, test_labels)