When using TestAssembler with polytomous items (GRM/GPCM), initialisation fails with ValueError: model has to be specified for polytomous items.
AdaptiveTest.init calls generate_response_pattern without forwarding the model parameter:
pythonself.item_pool.simulated_responses = generate_response_pattern(
ability=self.true_ability_level,
items=self.item_pool.test_items,
seed=kwargs["seed"] if "seed" in kwargs.keys() else None
# model is never passed
)
TestAssembler accepts model_type and **kwargs but neither reaches this call. Fix would be:
pythonmodel=kwargs["model"] if "model" in kwargs.keys() else None,
When using TestAssembler with polytomous items (GRM/GPCM), initialisation fails with ValueError: model has to be specified for polytomous items.
AdaptiveTest.init calls generate_response_pattern without forwarding the model parameter:
TestAssembler accepts model_type and **kwargs but neither reaches this call. Fix would be:
pythonmodel=kwargs["model"] if "model" in kwargs.keys() else None,