def test_find_top_spans_gap(self):
GAP = 0.0
v = [GAP, 0.9, 0.6, GAP, GAP, 1., GAP, GAP, GAP]
print(v)
topspans = find_top_spans(v, maxgap=1)
self.assertEqual(2, len(topspans))
topspans = find_top_spans(v, maxgap=3)
self.assertEqual(1, len(topspans))
def test_find_top_spans_limit(self):
v = [0., 0.9, 0.6, 0., 0., 0., 0., 0., 1.]
topspans = find_top_spans(v, limit=1)
print(topspans)
self.assertEqual(1, len(topspans))
sp: slice = topspans[0]
self.assertEqual(8, sp.start)
self.assertEqual(9, sp.stop)
def nn_get_tag_value(tagname: str,
textmap: TextMap,
semantic_map: DataFrame,
threshold=0.3) -> SemanticTag or None:
att = semantic_map[tagname].values
slices = find_top_spans(att, threshold=threshold,
limit=1) # TODO: estimate per-tag thresholds
if len(slices) > 0:
span = slices[0].start, slices[0].stop
value = textmap.text_range(span)
tag = SemanticTag(tagname, value, span)
tag.confidence = float(att[slices[0]].mean())
return tag
return None
def test_find_top_spans_no_resuslts(self):
v = [0.]
topspans = find_top_spans(v, limit=100)
print(topspans)
self.assertEqual(0, len(topspans))
def test_find_top_spans(self):
v = [0., 1., 0.6, 0., 0.]
topspans = find_top_spans(v, limit=1)
print(topspans)
self.assertEqual(1, len(topspans))
def test_find_top_spans_empty(self):
v = []
topspans = find_top_spans(v, limit=100)
print(topspans)
self.assertEqual(0, len(topspans))