sarsa_trace_update# sarsa_trace_update(model_free_values, second_stage_values, starting_state, action, reached_second_stage, reward, learning_rate, eligibility_trace)[source]# sarsa_trace, applied to the values in place. Parameters: model_free_valuesnumpy.ndarrayThe first-stage model-free values, shape (states, actions). Updated in place. second_stage_valuesnumpy.ndarrayThe second-stage values, a 1D array. Updated in place. starting_state, action, reached_second_stageint reward, learning_rate, eligibility_tracefloat Returns: stage1_prediction_error, stage2_prediction_errorfloat
