Mercurial > pylearn

comparison dataset.py @ 313:009ce84e9f52

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
behaviour is now the same as a list in pylearn, so if len(ds) = 10, ds[10] raise an IndexError, same thing for ds[[1,10]], and ds[0:14:1] returns 10 elements
author Thierry Bertin-Mahieux <bertinmt@iro.umontreal.ca>
date Wed, 11 Jun 2008 13:53:39 -0400
parents ebccfd05ccd5
children 105b54ac8260
comparison
equal deleted inserted replaced
312:96cca78de3ed 313:009ce84e9f52
453 the data are actually stored in a memory array. 453 the data are actually stored in a memory array.
454 """ 454 """
455 455
456 if type(i) is int: 456 if type(i) is int:
457 assert i >= 0 # TBM: see if someone complains and want negative i 457 assert i >= 0 # TBM: see if someone complains and want negative i
458 if i >= len(self) :
459 raise IndexError
458 i_batch = self.minibatches_nowrap(self.fieldNames(), 460 i_batch = self.minibatches_nowrap(self.fieldNames(),
459 minibatch_size=1, n_batches=1, offset=i) 461 minibatch_size=1, n_batches=1, offset=i)
460 return DataSet.MinibatchToSingleExampleIterator(i_batch).next() 462 return DataSet.MinibatchToSingleExampleIterator(i_batch).next()
461 463
462 #if i is a contiguous slice 464 #if i is a contiguous slice
463 if type(i) is slice and (i.step in (None, 1)): 465 if type(i) is slice and (i.step in (None, 1)):
464 offset = 0 if i.start is None else i.start 466 offset = 0 if i.start is None else i.start
465 upper_bound = len(self) if i.stop is None else i.stop 467 upper_bound = len(self) if i.stop is None else i.stop
468 upper_bound = min(len(self) , upper_bound)
466 #return MinibatchDataSet(self.minibatches_nowrap(self.fieldNames(), 469 #return MinibatchDataSet(self.minibatches_nowrap(self.fieldNames(),
467 # minibatch_size=upper_bound - offset, 470 # minibatch_size=upper_bound - offset,
468 # n_batches=1, 471 # n_batches=1,
469 # offset=offset).next()) 472 # offset=offset).next())
470 # now returns a LookupList 473 # now returns a LookupList
484 if hasattr(i, '__getitem__'): 487 if hasattr(i, '__getitem__'):
485 for idx in i: 488 for idx in i:
486 #dis-allow nested slices 489 #dis-allow nested slices
487 if not isinstance(idx, int): 490 if not isinstance(idx, int):
488 raise TypeError(idx) 491 raise TypeError(idx)
492 if idx >= len(self) :
493 raise IndexError
489 # call back into self.__getitem__ 494 # call back into self.__getitem__
490 examples = [self.minibatches_nowrap(self.fieldNames(), 495 examples = [self.minibatches_nowrap(self.fieldNames(),
491 minibatch_size=1, n_batches=1, offset=ii).next() 496 minibatch_size=1, n_batches=1, offset=ii).next()
492 for ii in i] 497 for ii in i]
493 # re-index the fields in each example by field instead of by example 498 # re-index the fields in each example by field instead of by example