e83bc935a8b701d1d38bd1dc2b8a3fd14a9cdecf,src/metamodels/crosscat.py,CrosscatMetamodel,row_column_predictive_probability,#CrosscatMetamodel#Any#Any#Any#Any#Any#,1158

Before Change

            value = row[0]
        if value is None:
            return None
        code = crosscat_value_to_code(bdb, generator_id, M_c, colno, value)
        cc_colno = crosscat_cc_colno(bdb, generator_id, colno)
        X_L_list = self._crosscat_latent_state(bdb, generator_id, modelno)
        X_D_list = self._crosscat_latent_data(bdb, generator_id, modelno)
        row_id, X_L_list, X_D_list = \
            self._crosscat_get_row(bdb, generator_id, rowid, X_L_list,
                X_D_list)
        r = self._crosscat.simple_predictive_probability_multistate(
            M_c=M_c,
            X_L_list=X_L_list,
            X_D_list=X_D_list,
            Y=[],
            Q=[(row_id, cc_colno, code)],
        )
        return math.exp(r)

After Change

        if value is None:
            return None
        r = self.logpdf_joint(
            bdb, generator_id, [(rowid, colno, value)], [], modelno)
        return math.exp(r)

    def predict_confidence(self, bdb, generator_id, modelno, colno, rowid,

Instances