def test_main(self):
tdm = build_hamlet_jz_corpus_with_meta()
features = FeatureLister(tdm._mX, tdm._metadata_idx_store,
tdm.get_num_docs()).output()
self.assertEqual(features, [{
'cat4': 2,
'cat3': 1
}, {
'cat4': 2
}, {
'cat5': 1,
'cat3': 2
}, {
'cat6': 2,
'cat9': 1
}, {
'cat4': 2,
'cat3': 1
}, {
'cat2': 1,
'cat1': 2
}, {
'cat2': 2,
'cat5': 1
}, {
'cat4': 1,
'cat3': 2
}])
def list_extra_features(self):
'''
Returns
-------
List of dicts. One dict for each document, keys are metadata, values are counts
'''
return FeatureLister(self._mX, self._metadata_idx_store,
self.get_num_docs()).output()