POMDPPlanners.tests.test_environments.test_nuplan_pomdp.test_nuplan_generative_models package

Submodules

POMDPPlanners.tests.test_environments.test_nuplan_pomdp.test_nuplan_generative_models.test_nuplan_factored_model_pomdp module

POMDPPlanners.tests.test_environments.test_nuplan_pomdp.test_nuplan_generative_models.test_nuplan_kinematic_model_pomdp module

POMDPPlanners.tests.test_environments.test_nuplan_pomdp.test_nuplan_generative_models.test_nuplan_model_pomdp module

Unit tests for the abstract nuPlan generative-model base (schema + observation composition).

The base is exercised through the concrete FactoredNuPlanModelPOMDP, which supplies the dynamics the abstract base leaves open while inheriting the shared schema and observation model.

POMDPPlanners.tests.test_environments.test_nuplan_pomdp.test_nuplan_generative_models.test_nuplan_model_pomdp.test_encode_observation_degrades_each_channel()[source]

encode_observation perceives every channel of the raw world reading.

Purpose: Validates the single raw-observation seam composes the channel models.

Given: A raw {ego, agents} observation with one distant agent When: encode_observation is applied Then: the perceived observation keeps both channels and gates the distant agent out

Test type: unit

Return type:: None

POMDPPlanners.tests.test_environments.test_nuplan_pomdp.test_nuplan_generative_models.test_nuplan_model_pomdp.test_get_actions_indexes_the_presets()[source]

The discrete action set enumerates the preset indices.

Purpose: Validates get_actions returns indices into action_presets.

Given: A factored nuPlan model with the default presets When: get_actions is called Then: it returns 0..len(presets)-1

Test type: unit

Return type:: None

POMDPPlanners.tests.test_environments.test_nuplan_pomdp.test_nuplan_generative_models.test_nuplan_model_pomdp.test_hash_and_equality_agree_on_equal_observations()[source]

Equal observations hash equally (planner belief-child indexing contract).

Purpose: Validates is_equal_observation implies equal hash_observation.

Given: Two structurally-equal {ego, agents} observations When: is_equal_observation and hash_observation are compared Then: they report equal and hash to the same key

Test type: unit

Return type:: None

POMDPPlanners.tests.test_environments.test_nuplan_pomdp.test_nuplan_generative_models.test_nuplan_model_pomdp.test_observation_log_probability_sums_channel_densities()[source]

The composed density is the sum of the per-channel log-densities.

Purpose: Validates observation_log_probability composes channels additively.

Given: A model and a clean-derived observation of a zero state When: observation_log_probability scores that observation Then: it equals the sum of the ego and agents channel densities

Test type: unit

Return type:: None