#@ s1: DakotaConfig=HAVE_ACRO
#@ s2: DakotaConfig=HAVE_ACRO
#@ s3: DakotaConfig=HAVE_ACRO
#@ s4: DakotaConfig=HAVE_ACRO
#@ s5: DakotaConfig=HAVE_ACRO

environment,
	  method_pointer = 'EPISTEMIC'

#################################
# begin EPISTEMIC specification #
#################################
method,
	id_method = 'EPISTEMIC'
	model_pointer = 'EPIST_M'
	global_interval_est lhs					#s0
#	global_interval_est sbo					#s1,#s2,#s3,#s4
	  samples = 10						#s0,#s1,#s2,#s3,#s4
#	  max_iterations = 10					#s1,#s2,#s3,#s4
#	global_interval_est ea					#s5
#	  max_iterations = 5					#s5
	  seed = 12347
	  output quiet

model,
	id_model = 'EPIST_M'
	nested
	  variables_pointer  = 'EPIST_V'
	  sub_method_pointer = 'ALEATORY'
	  responses_pointer  = 'EPIST_R'
	  primary_variable_mapping   = 'P'    'M'    'Y'    'ModelForm'	
	  secondary_variable_mapping = 'mean' 'mean' 'mean' ''
#	  primary_response_mapping   = 0. 0. 1.
	  primary_response_mapping   = 1. 0. 0.
	  			       0. 1. 0.

variables,
	id_variables = 'EPIST_V'
	continuous_interval_uncertain = 3
	  lower_bounds = 400 1750 4
	  upper_bounds = 600 2250 6
	  descriptors = 'P_mean' 'M_mean' 'Y_mean'		
	discrete_uncertain_set
	  integer = 1
	    set_values =  1   2   3   4
#	    set_values =  1   2   4         # model 3 has most nonlinear discrep
#	    set_values =  1   2		  # models 1,2 are most similar
	    descriptors = 'EpistModelForm'		

responses,
	id_responses = 'EPIST_R'
	response_functions = 2
	response_descriptors = 'mean_limit' 'std_dev_limit'
#	response_functions = 1
#	response_descriptors = 'cdf_beta'
	no_gradients
	no_hessians

################################
# begin ALEATORY specification #
################################
method,
	id_method = 'ALEATORY'
	model_pointer = 'ALEAT_M'
	polynomial_chaos askey			#s0,#s1,#s5
#	polynomial_chaos 			#s2
#	stoch_collocation askey			#s3
#	stoch_collocation 			#s4
	  sparse_grid_level = 2
#	  num_response_levels = 0 1
	  response_levels = 0.0
          compute reliabilities
          cumulative distribution
	  output silent

model,
	id_model = 'ALEAT_M'
	single
	  variables_pointer = 'ALEAT_V'
	  interface_pointer = 'ALEAT_I'
	  responses_pointer = 'ALEAT_R'

variables,
	id_variables = 'ALEAT_V'
	normal_uncertain = 2
	  means             =  500.0   2000.0
	  std_deviations    =  100.0    400.0
	  descriptors       =   'P'      'M'
	lognormal_uncertain = 1
       	  means             =  5.0
	  std_deviations    =  0.5
	  descriptors       =  'Y'
        uniform_uncertain = 2
	  lower_bounds      5.    15.
	  upper_bounds     15.    25.
          descriptors      'b'    'h'
	uncertain_correlation_matrix =  1   0.5 0 0 0
					0.5 1   0 0 0
					0   0   1 0 0
					0   0   0 1 0
					0   0   0 0 1
	discrete_state_set
	  integer = 1
	    set_values     = 1   2   3   4
	    descriptors    = 'ModelForm'	

interface,
	id_interface = 'ALEAT_I'
	direct
	  analysis_driver = 'mf_short_column'
#	  deactivate evaluation_cache restart_file

responses,
	id_responses = 'ALEAT_R'
#	response_descriptors = 'area' 'limit_state'
#	response_functions = 2
	response_descriptors = 'limit_state'
	response_functions = 1
	no_gradients
	no_hessians