def __init__(self, *args, **_): GeneralizedPairState.__init__(self, *args) self.dp = config.preparers[config.preparer_index][0](constants.window_size) self.clf_fname = 'data/clf/{}{}{}.clf'.format( PairClassifier.get_name(), config.preparers[config.preparer_index][2], constants.window_size, ) self.clf = self._get_classifier() self.annotations, self.ann_x, self.ann_y = None, None, None self.emission_table = None
def __init__(self, *args, **_): GeneralizedPairState.__init__(self, *args) self.dp = config.preparers[config.preparer_index][0]( constants.window_size) self.clf_fname = 'data/clf/{}{}{}.clf'.format( PairClassifier.get_name(), config.preparers[config.preparer_index][2], constants.window_size, ) self.clf = self._get_classifier() self.annotations, self.ann_x, self.ann_y = None, None, None self.emission_table = None
def __init__(self, *args, **kwargs): ClassifierState.__init__(self, *args, **kwargs) self.dp = self._get_preparer(0) if config.same_classifier: suffix = '' else: suffix = '_indel' self.clf_fname = 'data/clf/{}{}{}{}.clf'.format( PairClassifier.get_name(), config.preparers[config.preparer_index][2], constants.window_size, suffix, ) self.clf = self._get_classifier()
data = ( ( seq_x, pos, create_annotation(ann_x), seq_y, pos, create_annotation(ann_y), ) for seq_x in itertools.product('ACGT', repeat=window_size) for seq_y in itertools.product('ACGT', repeat=window_size) for ann_x in itertools.product('01', repeat=window_size) for ann_y in itertools.product('01', repeat=window_size) ) else: data = ( ( seq_x, pos, create_annotation(ann_x), seq_y, pos, create_annotation(ann_y), ) for seq_x in 'ACGT' for seq_y in 'ACGT' for ann_x in '01' for ann_y in '01' ) s = sum(clf.multi_prepare_predict(data)) return s if __name__ == '__main__': dp = DataPreparer(window_size) clf = PairClassifier(dp, filename='data/randomforest1.clf') print('Normalization constant:', compute_norm_constant(clf))
def _get_classifier(self): return PairClassifier(self.dp, filename=self.clf_fname, inverted=config.same_classifier, use_global_classifier=config.same_classifier)