本文整理汇总了Python中mvpa2.datasets.base.Dataset.sa['targets']方法的典型用法代码示例。如果您正苦于以下问题：Python Dataset.sa['targets']方法的具体用法？Python Dataset.sa['targets']怎么用？Python Dataset.sa['targets']使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类mvpa2.datasets.base.Dataset的用法示例。
在下文中一共展示了Dataset.sa['targets']方法的2个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: create_mvpa_dataset
# 需要导入模块: from mvpa2.datasets.base import Dataset [as 别名]
# 或者: from mvpa2.datasets.base.Dataset import sa['targets'] [as 别名]
def create_mvpa_dataset(aXData1, aXData2, chunks, labels):
feat_list = []
for x1, x2, chunk in zip(aXData1, aXData2, chunks):
feat_list.append([x1, x2])
data = Dataset(samples=feat_list)
data.sa['id'] = range(0,len(labels))
data.sa['chunks'] = chunks
data.sa['targets'] = labels
return data示例2: _sl_call
# 需要导入模块: from mvpa2.datasets.base import Dataset [as 别名]
# 或者: from mvpa2.datasets.base.Dataset import sa['targets'] [as 别名]
#.........这里部分代码省略.........
"roi's are expensive at this point. Get them from the "
".ca value of the original instance before "
"calling again and using reuse_neighbors")
else:
raise RuntimeError("Should not be reachable")
# Since this is ad-hoc implementation of the searchlight, we are not passing
# those via ds.a but rather assign directly to self.ca
self.ca.roi_sizes = roi_sizes
indexsum = self._indexsum
if indexsum == 'sparse':
if not self.reuse_neighbors or self.__roi_fids is None:
if __debug__:
debug('SLC',
'Phase 4b. Converting neighbors to sparse matrix '
'representation')
# convert to "sparse representation" where column j contains
# 1s only at the roi_fids[j] indices
roi_fids = inds_to_coo(roi_fids,
shape=(dataset.nfeatures, nroi_fids))
indexsum_fx = lastdim_columnsums_spmatrix
elif indexsum == 'fancy':
indexsum_fx = lastdim_columnsums_fancy_indexing
else:
raise ValueError, \
"Do not know how to deal with indexsum=%s" % indexsum
# Store roi_fids
if self.reuse_neighbors and self.__roi_fids is None:
self.__roi_fids = roi_fids
# 5. Lets do actual "splitting" and "classification"
if __debug__:
debug('SLC', 'Phase 5. Major loop' )
for isplit, split in enumerate(splits):
if __debug__:
debug('SLC', ' Split %i out of %i' % (isplit, nsplits))
# figure out for a given splits the blocks we want to work
# with
# sample_indicies
training_sis = split[0].samples[:, 0]
testing_sis = split[1].samples[:, 0]
# That is the GNB specificity
targets, predictions = self._sl_call_on_a_split(
split, X, # X2 might light to go
training_sis, testing_sis,
## training_nsamples, # GO? == np.sum(pl.nsamples)
## training_non0labels,
## pl.sums, pl.means, pl.sums2, pl.variances,
# passing nroi_fids as well since in 'sparse' way it has no 'length'
nroi_fids, roi_fids,
indexsum_fx,
labels_numeric,
)
# assess the errors
if __debug__:
debug('SLC', " Assessing accuracies")
if errorfx is mean_mismatch_error:
results[isplit, :] = \
(predictions != targets[:, None]).sum(axis=0) \
/ float(len(targets))
all_cvfolds += [isplit]
elif errorfx:
# somewhat silly but a way which allows to use pre-crafted
# error functions without a chance to screw up
results.append(
np.array([errorfx(fpredictions, targets)
for fpredictions in predictions.T]))
all_cvfolds += [isplit] * len(targets)
else:
# and if no errorfx -- we just need to assign original
# labels to the predictions BUT keep in mind that it is a matrix
results.append(assign_ulabels(predictions))
all_targets += [ulabels[i] for i in targets]
all_cvfolds += [isplit] * len(targets)
pass # end of the split loop
if isinstance(results, list):
# we have just collected them, now they need to be vstacked
results = np.vstack(results)
assert(results.ndim >= 2)
if __debug__:
debug('SLC', "%s._call() is done in %.3g sec" %
(self.__class__.__name__, time.time() - time_start))
out = Dataset(results)
if all_targets:
out.sa['targets'] = all_targets
out.sa['cvfolds'] = all_cvfolds
out.fa['center_ids'] = roi_ids
return out