diff options
| -rw-r--r-- | lib/parameters.py | 26 |
1 files changed, 14 insertions, 12 deletions
diff --git a/lib/parameters.py b/lib/parameters.py index 603e56d..ffbbe70 100644 --- a/lib/parameters.py +++ b/lib/parameters.py @@ -574,7 +574,7 @@ class ModelAttribute: # In this case, only one of them must be used for parameter-dependent model attribute detection and modeling self.codependent_param_pair = codependent_param self.codependent_params = [list() for x in self.log_param_names] - self.ignore_param = dict() + self.ignore_codependent_param = dict() # Static model used as lower bound of model accuracy if data is not None: @@ -802,7 +802,7 @@ class ModelAttribute: param1_numeric_count >= param2_numeric_count and params_are_pairwise_none ): - self.ignore_param[param2_index] = True + self.ignore_codependent_param[param2_index] = True self.codependent_params[param1_index].append(param2_index) logger.debug( f"{self.name} {self.attr}: parameters ({self.log_param_names[param1_index]}, {self.log_param_names[param2_index]}) are codependent. Ignoring {self.log_param_names[param2_index]}" @@ -811,7 +811,7 @@ class ModelAttribute: param2_numeric_count >= param1_numeric_count and params_are_pairwise_none ): - self.ignore_param[param1_index] = True + self.ignore_codependent_param[param1_index] = True self.codependent_params[param2_index].append(param1_index) logger.debug( f"{self.name} {self.attr}: parameters ({self.log_param_names[param1_index]}, {self.log_param_names[param2_index]}) are codependent. Ignoring {self.log_param_names[param1_index]}" @@ -864,7 +864,7 @@ class ModelAttribute: for param_index, param_name in enumerate(self.param_names): if ( self.stats.depends_on_param(param_name) - and not param_index in self.ignore_param + and not param_index in self.ignore_codependent_param ): return True return False @@ -873,7 +873,7 @@ class ModelAttribute: for param_index, param_name in enumerate(self.param_names): if ( self.stats.depends_on_param(param_name) - and not param_index in self.ignore_param + and not param_index in self.ignore_codependent_param ): param_values = list(map(lambda x: x[param_index], self.param_values)) if not all(map(lambda n: n is None or is_numeric(n), param_values)): @@ -891,7 +891,7 @@ class ModelAttribute: for param_index, param_name in enumerate(self.param_names): if ( self.stats.depends_on_param(param_name) - and not param_index in self.ignore_param + and not param_index in self.ignore_codependent_param ): by_param = self._by_param_for_index(param_index) ret.append( @@ -907,7 +907,7 @@ class ModelAttribute: param_index = len(self.param_names) + arg_index if ( self.stats.depends_on_arg(arg_index) - and not param_index in self.ignore_param + and not param_index in self.ignore_codependent_param ): by_param = self._by_param_for_index(param_index) ret.append( @@ -1215,7 +1215,7 @@ class ModelAttribute: ffs_eligible_params = list() ffs_unsuitable_params = list() for param_index in range(param_count): - if param_index in self.ignore_param: + if param_index in self.ignore_codependent_param: continue unique_values = list(set(map(lambda p: p[param_index], parameters))) if None in unique_values: @@ -1229,7 +1229,7 @@ class ModelAttribute: ffs_unsuitable_params.append(param_index) for param_index in range(param_count): - if param_index in self.ignore_param: + if param_index in self.ignore_codependent_param: loss.append(np.inf) continue @@ -1260,7 +1260,9 @@ class ModelAttribute: std_by_param = _mean_std_by_params( by_param, distinct_values_by_param_index, - list(self.ignore_param.keys()) + irrelevant_params + [param_index], + list(self.ignore_codependent_param.keys()) + + irrelevant_params + + [param_index], ) if not _depends_on_param( None, std_by_param, std_lut, relevance_threshold @@ -1293,7 +1295,7 @@ class ModelAttribute: child_data_by_scalar = partition_by_param( child_data, child_param, - ignore_parameters=list(self.ignore_param.keys()) + ignore_parameters=list(self.ignore_codependent_param.keys()) + ffs_unsuitable_params, ) logger.debug(f"got {len(child_data_by_scalar)} partitions") @@ -1335,7 +1337,7 @@ class ModelAttribute: data_by_scalar = partition_by_param( data, parameters, - ignore_parameters=list(self.ignore_param.keys()) + ignore_parameters=list(self.ignore_codependent_param.keys()) + ffs_unsuitable_params, ) if np.all( |