def __call__(self, dataset):
# create the dissimilarity matrix for the data in the input dataset
self.dset_dsm = DSMatrix(dataset.samples, self.dset_metric)
in_vec = self.dsmatrix.getVectorForm()
dset_vec = self.dset_dsm.getVectorForm()
# concatenate the two vectors, send to dissimlarity function
test_mat = N.asarray([in_vec, dset_vec])
test_dsmatrix = DSMatrix(test_mat, self.output_metric)
# return correct dissimilarity value
return test_dsmatrix.getFullMatrix()[0, 1]
class DSMDatasetMeasure(DatasetMeasure):
"""DSMDatasetMeasure creates a DatasetMeasure object
where metric can be one of 'euclidean', 'spearman', 'pearson'
or 'confusion'"""
def __init__(self, dsmatrix, dset_metric, output_metric='spearman'):
DatasetMeasure.__init__(self)
self.dsmatrix = dsmatrix
self.dset_metric = dset_metric
self.output_metric = output_metric
self.dset_dsm = []
def __call__(self, dataset):
# create the dissimilarity matrix for the data in the input dataset
self.dset_dsm = DSMatrix(dataset.samples, self.dset_metric)
in_vec = self.dsmatrix.getVectorForm()
dset_vec = self.dset_dsm.getVectorForm()
# concatenate the two vectors, send to dissimlarity function
test_mat = N.asarray([in_vec, dset_vec])
test_dsmatrix = DSMatrix(test_mat, self.output_metric)
# return correct dissimilarity value
return test_dsmatrix.getFullMatrix()[0, 1]