Use Assignment Expression (Walrus) In Conditional

sigoptlite/builders.py

@@ -254,8 +254,7 @@ def validate_input_dict(cls, input_dict):
   def create_object(cls, **input_dict):
     cls.set_object(input_dict, "assignments", LocalAssignments)
     cls.set_list_of_objects(input_dict, field="values", local_class=MetricEvaluationBuilder)
-    values = input_dict.get("values")
-    if values:
+    if values := input_dict.get("values"):
       input_dict["metric_evaluations"] = {me.name: me for me in values}
     input_dict.pop("values", None)
     cls.set_object(input_dict, "task", LocalTaskBuilder)
@@ -273,8 +272,7 @@ class MetricEvaluationBuilder(BuilderBase):
   def validate_input_dict(cls, input_dict):
     assert isinstance(input_dict["name"], str)
     assert isinstance(input_dict["value"], (int, float))
-    value_stddev = input_dict.get("value_stddev", None)
-    if value_stddev is not None:
+    if (value_stddev := input_dict.get("value_stddev", None)) is not None:
       assert isinstance(input_dict["value_stddev"], (int, float))
       assert input_dict["value_stddev"] >= 0
       assert set(input_dict.keys()) == {"name", "value", "value_stddev"}

sigoptlite/driver.py

@@ -112,7 +112,6 @@ def path_to_route(self, path, method):
 
   def request(self, method, path, data, headers):
     route = self.path_to_route(path, method)
-    handler = self.routes.get(route, {}).get(method)
-    if handler is None:
+    if (handler := self.routes.get(route, {}).get(method)) is None:
       raise Exception(f"{PRODUCT_NAME} only supports the following routes: {' '.join(self.routes.keys())}")
     return handler(data)

sigoptlite/models.py

@@ -254,8 +254,7 @@ def get_optimized_measurements_for_maximization(self, experiment):
     return [self.get_value_for_maximization(metric) for metric in experiment.optimized_metrics]
 
   def get_value_for_maximization(self, metric):
-    value = self.get_metric_evaluation_by_name(metric.name).value
-    if value is None:
+    if (value := self.get_metric_evaluation_by_name(metric.name).value) is None:
       raise Exception(f"Metric `{metric.name}` is not in observation data.")
     if metric.is_minimized:
       return -value

sigoptlite/sources.py

@@ -291,8 +291,7 @@ def make_assignments_from_point(experiment, point):
   def get_point_from_assignments(experiment, assignments):
     point = numpy.empty(experiment.dimension)
     for i, parameter in enumerate(experiment.parameters):
-      parameter_value = assignments.get(parameter.name, None)
-      if parameter_value is None:
+      if (parameter_value := assignments.get(parameter.name, None)) is None:
         parameter_value = replacement_value_if_missing(parameter)
       if parameter.has_log_transformation:
         parameter_value = numpy.log10(parameter_value)

sigoptlite/validators.py

@@ -204,8 +204,7 @@ def validate_experiment(experiment, cls_name):
   if not experiment.parallel_bandwidth == 1:
     raise ValueError(f"{cls_name} must have parallel_bandwidth == 1")
 
-  observation_budget = experiment.observation_budget
-  if observation_budget is None:
+  if (observation_budget := experiment.observation_budget) is None:
     if experiment.num_solutions > 1:
       raise ValueError(f"observation_budget is required for a {cls_name} with multiple solutions")
     if experiment.requires_pareto_frontier_optimization:
@@ -250,8 +249,7 @@ def validate_experiment(experiment, cls_name):
     validate_conditionals_for_experiment(experiment)
 
   # Check feature viability of multitask
-  tasks = experiment.tasks
-  if tasks:
+  if tasks := experiment.tasks:
     if experiment.requires_pareto_frontier_optimization:
       raise ValueError(f"{cls_name} cannot have both tasks and multiple optimized metrics")
     if experiment.has_constraint_metrics:
@@ -441,8 +439,7 @@ def validate_constraints_for_experiment(experiment):
 
     term_types = []
     for term in terms:
-      coeff = term.weight
-      if coeff == 0:
+      if (coeff := term.weight) == 0:
         continue
       name = term.name
       if name in integer_params_names: