0

I have two separate folders containing 3D arrays (data), each folder contains files of the same classification. I used mxnet.gluon.data.ArrayDataset() create datasets for each label respectively. Is there a way to combine these two datasets into the final training dataset that combines both classifications? The new data sets are different size.

e.g

A_data = mx.gluon.data.ArrayDataset(list2,label_A )
noA_data = mx.gluon.data.ArrayDataset(list,label_noA)

^ I want to combine A_data and noA_data for a complete dataset.

Additionally, is there an easier way to combine the two folders with its classification into a mxnet dataset from the get-go? That would also solve my problem.

accAscrub
  • 609
  • 1
  • 6
  • 11

1 Answers1

0

You could create an ArrayDataset that contains both, if list and list2 are both python lists then you could do something like

full_data = mx.gluon.data.dataset.ArrayDataset(list + list2, label_noA + labelA)

where len(label_noA) == len(list) and len(label_A) == len(list2)

sad-
  • 101
  • 2