1Blurb:: 2Experimental capability to generate a random field representation. 3from data, from simulation 4runs, or from a covariance matrix. The representation may then be sampled 5for use as a random field input to another simulation. THIS IS AN EXPERIMENTAL CAPABILITY. 6 7Description:: 8Capability to generate a random field representation from data, from simulation 9runs, or from a covariance matrix. The random field may then be sampled 10for use as a random field input to another simulation. THIS IS AN EXPERIMENTAL CAPABILITY UNDER ACTIVE DEVELOPMENT. 11 12<b> Default Behavior </b> 13There are three main sections of the \c random_field model. 14The first section tells Dakota what data to use to build the random field. 15This is specified with \c build_source. The source of data to build the 16random field may be a file with data (where the N rows of data correspond to 17N samples of the random field and the M columns correspond to field values), 18or it may be a simulation that generates field data, or it may be specified 19given a mesh and a covariance matrix governing how the field varies over the mesh. 20In the case of using a simulation to generate field data, the simulation is 21defined with \c dace_method_pointer. In the case of using a mesh and a 22covariance, the form of the covariance is defined with \c analytic_covariance. 23 24The next section of the random fiel model specifies the form of the expansion, 25\c expansion_form. This can be either a Karhunen-Loeve expansion or a 26Principal components analysis. These are very similar: both involve the eigenvalues 27of the covariance matrix of the field data. The only difference is in the treatment 28of the estimation of the coefficients of the eigenvector basis functions. In the 29PCA case, we have developed an approach which makes the coefficients explicit 30functions of the uncertain variables used to generate the random field. 31The specification of the random field can also include the number of bases 32to retain or a truncation tolerance, which defines the percent variance 33that the expansion should capture. 34 35The final section of the random field model allows the user to specify 36a pointer to a model over which the random field will be propagated, 37meaning the model which will be driven with the random field input. 38This part of the specification is optional: one can build a random field but 39not use it in a downstream model. 40 41 42Topics:: 43 44Examples:: 45As stated above, this is an emerging capability. 46The syntax currently looks like the following: 47\verbatim 48random_field 49 build_source 50 rf_data_file | dace_method_pointer | analytic_covariance 51 expansion_form 52 karhunen_loeve | principal_components 53 expansion_bases 54 truncation_tolerance 55 propagation_model_pointer 56\endverbatim 57 58 59Theory:: 60Faq:: 61See_Also:: 62