Changeset 212

12/11/10 09:22:20 (3 years ago)

Changes to add more description to the response mappings
for nested models and update the Bayesian class on the main
Developer's page.

2 edited


  • core/trunk/docs/Dev_Main.dox

    r169 r212  
    183183Dakota::NonDInterval "NonDInterval" (interval-based epistemic 
    184184methods).  Bayesian calibration methods are prototyped in \ref 
    185 Dakota::NonDBayesCal "NonDBayesCal". 
     185Dakota::NonDBayesCalibration "NonDBayesCalibration".  At this point, we have an initial implementation of the LANL GPMSA code in \ref Dakota::NonDGPMSABayesCalibration "NonDGPMSABayesianCalibration". 
    187187<ul> <li> \ref Dakota::NonDSampling "NonDSampling" is further 
  • core/trunk/docs/Ref_Model.dox

    r204 r212  
    979979For the response mappings, the primary and secondary specifications 
    980980provide real-valued multipliers to be applied to sub-iterator response 
    981 results. The sub-iterator response results are defined as follows for 
     981results. The idea is that the responses from the inner loop are mapped to the outer loop.  For example, if the nested model is an uncertainty quantification, aleatory statistics from the inner loop such as the mean inner loop response are mapped to the outer level, where they are treated epistemically, so that intervals on the mean (for example) are calculated. The response mapping defines a vector which multiplies the values from the inner loop to the outer loop.  Each row of the mapping corresponds to one outer loop response, where each column of the mapping corresponds to a value from the inner loop.  Depending on the number of responses and the particular attributes calculated on the inner loop, there will be a vector of inner loop response values that need to be accounted for in the mapping.  The sub-iterator response results are defined as follows for 
    982982different sub-iterator types: 
    10251025the latter two of which are inserted into the mean distribution 
    10261026parameters of sub-model variables \c 'X' and \c 'Y' (option 1 above). 
     1027In this particular example, there are 9 inner loop response attributes and 3 outer loop response functions (one primary response function and 2 secondary functions, such as one objective and two constraints).  Each row of the response mapping is a vector which is multiplied (e.g. dot-product) against the 9 sub-iterator values to determine the outer loop function.  For example, the primary response mapping only picks up the first value from the inner loop.  This first value is the mean of the first response function on the inner loop.  
    10271028The response mappings correspond to 9 sub-iterator response functions 
    1028 (e.g., a set of UQ final statistics for 3 response functions, each 
    1029 with a mean, a standard deviation, and a level mapping).  The primary 
     1029(e.g., there are 3 response functions, each with a set of UQ final statistics:  each has a mean, a standard deviation, and a level mapping in this example.  If no probability or reliability levels are specified, the responses would only have a mean and standard deviation).  The primary 
    10301030response mapping maps the first sub-iterator response function (mean) 
    10311031into a single objective function, least squares term, or generic 
    10381038standard deviations) into another top-level nonlinear constraint 
    10391039(these top-level nonlinear constraints may be inequality or equality, 
    1040 as dictated by the top-level response specification). 
     1040as dictated by the top-level response specification). Note that in many cases, each particular sub-iterator response will be mapped to a unique outer loop response (for example, in the nested UQ case where one wants to determine an interval on each inner loop statistic).  In these cases, the response mapping will be the identity matrix.  That is, the primary response mapping will have N rows and N columns corresponding to the N sub-iterator response values, and the mapping matrix will have a value of one along the diagonal and zeros elsewhere. 
    10421042Table \ref T6d10 "6.10" provides the specification detail for the model 
Note: See TracChangeset for help on using the changeset viewer.