1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
|
from Orange.base import RandomForestModel, Learner
from Orange.classification import RandomForestLearner as RFClassification
from Orange.data import Variable
from Orange.modelling import SklFitter
from Orange.preprocess.score import LearnerScorer
from Orange.regression import RandomForestRegressionLearner as RFRegression
__all__ = ['RandomForestLearner']
class _FeatureScorerMixin(LearnerScorer):
feature_type = Variable
class_type = Variable
def score(self, data):
model = self.get_learner(data)(data)
return model.skl_model.feature_importances_, model.domain.attributes
class RandomForestLearner(SklFitter, _FeatureScorerMixin):
name = 'random forest'
__fits__ = {'classification': RFClassification,
'regression': RFRegression}
__returns__ = RandomForestModel
@property
def fitted_parameters(self) -> list[Learner.FittedParameter]:
return [self.FittedParameter("n_estimators", "Number of trees",
int, 1, None)]
|
