Downloads: 1
Diff of
/src/features/token_length_feature.py
[000000]
..
[735bb5]
Switch to unified view
| a | b/src/features/token_length_feature.py | ||
|---|---|---|---|
| 1 | # Base Dependencies |
||
| 2 | # ---------------- |
||
| 3 | import numpy |
||
| 4 | |||
| 5 | # Local Dependencies |
||
| 6 | # ------------------ |
||
| 7 | from models import RelationCollection |
||
| 8 | |||
| 9 | # 3rd-Party Dependencies |
||
| 10 | # ---------------------- |
||
| 11 | from sklearn.base import BaseEstimator |
||
| 12 | |||
| 13 | |||
| 14 | class TokenLengthFeature(BaseEstimator): |
||
| 15 | """ |
||
| 16 | TokenLengthFeature |
||
| 17 | |||
| 18 | Computes the number of tokens of each relation. This is used in evaluation. |
||
| 19 | """ |
||
| 20 | |||
| 21 | def __init__(self): |
||
| 22 | pass |
||
| 23 | |||
| 24 | def get_feature_names(self, input_features=None): |
||
| 25 | return ["token_length"] |
||
| 26 | |||
| 27 | def create_token_length_feature( |
||
| 28 | self, collection: RelationCollection |
||
| 29 | ) -> numpy.array: |
||
| 30 | features = [] |
||
| 31 | for doc in collection.tokens: |
||
| 32 | features.append([len(doc)]) |
||
| 33 | |||
| 34 | return numpy.array(features) |
||
| 35 | |||
| 36 | def fit(self, x: RelationCollection, y=None): |
||
| 37 | return self |
||
| 38 | |||
| 39 | def transform(self, x: RelationCollection, y=None) -> numpy.array: |
||
| 40 | return self.create_token_length_feature(x) |
||
| 41 | |||
| 42 | def fit_transform(self, x: RelationCollection, y=None) -> numpy.array: |
||
| 43 | return self.create_token_length_feature(x) |