Mercurial > ift6266

annotate data_generation/mnist_resized/rescale_mnist.py @ 647:47af8a002530 tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
changed Theano to ift6266 and remove numpy as we do not use code from numpy in this repository
author Razvan Pascanu <r.pascanu@gmail.com>
date Wed, 17 Oct 2012 09:26:14 -0400
parents 128bc92897f2
children
rev   line source
625
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
1 import numpy,cPickle,gzip,Image,pdb,sys
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
2
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
3
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
4 def zeropad(vect,img_size=(28,28),out_size=(32,32)):
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
5 delta = (numpy.abs(img_size[0]-out_size[0])/2,numpy.abs(img_size[1]-out_size[1])/2)
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
6 newvect = numpy.zeros(out_size)
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
7 newvect[delta[0]:-delta[0],delta[1]:-delta[1]] = vect.reshape(img_size)
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
8 return newvect.flatten()
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
9
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
10 def rescale(vect,img_size=(28,28),out_size=(32,32), filter=Image.NEAREST):
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
11 im = Image.fromarray(numpy.asarray(vect.reshape(img_size)*255.,dtype='uint8'))
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
12 return (numpy.asarray(im.resize(out_size,filter),dtype='float32')/255.).flatten()
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
13
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
14
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
15 #pdb.set_trace()
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
16 def rescale_mnist(newsize=(32,32),output_file='mnist_rescaled_32_32.pkl',mnist=cPickle.load(gzip.open('mnist.pkl.gz'))):
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
17 newmnist = []
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
18 for set in mnist:
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
19 newset=numpy.zeros((len(set[0]),newsize[0]*newsize[1]))
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
20 for i in xrange(len(set[0])):
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
21 print i,
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
22 sys.stdout.flush()
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
23 newset[i] = rescale(set[0][i])
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
24 newmnist.append((newset,set[1]))
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
25 cPickle.dump(newmnist,open(output_file,'w'),protocol=-1)
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
26 print 'Done rescaling'
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
27
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
28
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
29 def zeropad_mnist(newsize=(32,32),output_file='mnist_zeropadded_32_32.pkl',mnist=cPickle.load(gzip.open('mnist.pkl.gz'))):
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
30 newmnist = []
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
31 for set in mnist:
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
32 newset=numpy.zeros((len(set[0]),newsize[0]*newsize[1]))
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
33 for i in xrange(len(set[0])):
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
34 print i,
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
35 sys.stdout.flush()
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
36 newset[i] = zeropad(set[0][i])
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
37 newmnist.append((newset,set[1]))
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
38 cPickle.dump(newmnist,open(output_file,'w'),protocol=-1)
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
39 print 'Done padding'
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
40
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
41 if __name__ =='__main__':
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
42 print 'Creating resized datasets'
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
43 mnist_ds = cPickle.load(gzip.open('mnist.pkl.gz'))
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
44 #zeropad_mnist(mnist=mnist_ds)
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
45 rescale_mnist(mnist=mnist_ds)
128bc92897f2 Script to generate resized version of mnist in two different ways: rescaling and zeropadding
Salah Rifai <salahmeister@gmail.com>
parents:
diff changeset
46 print 'Finished.'