Mercurial > pylearn

comparison pylearn/datasets/shapeset1.py @ 617:5120bf7c4694

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
More complete version of shapeset1 dataset.
author lamblinp@ip03.m
date Sat, 17 Jan 2009 19:06:21 -0500
parents 16f91ca016b1
children 8aef46b42cb5
comparison
equal deleted inserted replaced
616:d0f7a6f87adc 617:5120bf7c4694
7 import os 7 import os
8 import numpy 8 import numpy
9 9
10 from ..io.amat import AMat 10 from ..io.amat import AMat
11 from .config import data_root 11 from .config import data_root
12 from .dataset import Dataset
12 13
13 def _head(path, n): 14 def _head(path, n):
14 dat = AMat(path=path, head=n) 15 dat = AMat(path=path, head=n)
15 16
16 try: 17 try:
23 24
24 25
25 def head_train(n=10000): 26 def head_train(n=10000):
26 """Load the first Shapeset1 training examples. 27 """Load the first Shapeset1 training examples.
27 28
28 Returns two matrices: x, y. x has N rows of 1024 columns. Each row of x represents the 29 Returns two matrices: x, y.
29 32x32 grey-scale pixels in raster order. y is a vector of N integers. Each element y[i] 30 x has N rows of 1024 columns.
30 is the label of the i'th row of x. 31 Each row of x represents the 32x32 grey-scale pixels in raster order.
31 32 y is a vector of N integers between 0 and 2.
33 Each element y[i] is the label of the i'th row of x.
32 """ 34 """
33 path = os.path.join(data_root(), 'shapeset1','shapeset1_1cspo_2_3.10000.train.shape.amat') 35 path = os.path.join(data_root(), 'shapeset1','shapeset1_1cspo_2_3.10000.train.shape.amat')
34 return _head(path, n) 36 return _head(path, n)
35 37
36 def head_valid(n=5000): 38 def head_valid(n=5000):
37 """Load the first Shapeset1 validation examples. 39 """Load the first Shapeset1 validation examples.
38 40
39 Returns two matrices: x, y. x has N rows of 1024 columns. Each row of x represents the 41 Returns two matrices: x, y.
40 32x32 grey-scale pixels in raster order. y is a vector of N integers. Each element y[i] 42 x has N rows of 1024 columns.
41 is the label of the i'th row of x. 43 Each row of x represents the 32x32 grey-scale pixels in raster order.
42 44 y is a vector of N integers between 0 and 2.
45 Each element y[i] is the label of the i'th row of x.
43 """ 46 """
44 path = os.path.join(data_root(), 'shapeset1','shapeset1_1cspo_2_3.5000.valid.shape.amat') 47 path = os.path.join(data_root(), 'shapeset1','shapeset1_1cspo_2_3.5000.valid.shape.amat')
45 return _head(path, n) 48 return _head(path, n)
46 49
47 def head_test(n=5000): 50 def head_test(n=5000):
48 """Load the first Shapeset1 testing examples. 51 """Load the first Shapeset1 testing examples.
49 52
50 Returns two matrices: x, y. x has N rows of 1024 columns. Each row of x represents the 53 Returns two matrices: x, y.
51 32x32 grey-scale pixels in raster order. y is a vector of N integers. Each element y[i] 54 x has N rows of 1024 columns.
52 is the label of the i'th row of x. 55 Each row of x represents the 32x32 grey-scale pixels in raster order.
53 56 y is a vector of N integers between 0 and 2.
57 Each element y[i] is the label of the i'th row of x.
54 """ 58 """
55 path = os.path.join(data_root(), 'shapeset1','shapeset1_1cspo_2_3.5000.test.shape.amat') 59 path = os.path.join(data_root(), 'shapeset1','shapeset1_1cspo_2_3.5000.test.shape.amat')
56 return _head(path, n) 60 return _head(path, n)
57 61
58 def train_valid_test(ntrain=10000, nvalid=5000, ntest=5000): 62 def train_valid_test(ntrain=10000, nvalid=5000, ntest=5000):
59 return head_train(n=ntrain), head_valid(n=nvalid), head_test(n=ntest) 63 train_x, train_y = head_train(n=ntrain)
64 valid_x, valid_y = head_valid(n=nvalid)
65 test_x, test_y = head_test(n=test)
66
67 rval = Dataset()
68 rval.train = Dataset.Obj(x = train_x, y = train_y)
69 rval.valid = Dataset.Obj(x = valid_x, y = valid_y)
70 rval.test = Dataset.Obj(x = test_x, y = test_y)
71
72 rval.n_classes = 3
73 rval.img_shape = (32, 32)
74
75 return rval
60 76
61 77