www.delorie.com/gnu/docs/mcsim/mcsim_41.html   search  
 
Buy GNU books!


MCSim User' Manual

[ < ] [ > ]   [ << ] [ Up ] [ >> ]         [Top] [Contents] [Index] [ ? ]

6.4.5 Data() specification

Experimental observations of model variables, inputs, outputs, or parameters, can be specified with the Data() command. Markov chain Monte Carlo sampling requires that you specify Data() statements (see section 6.3.6 MCMC() specification; see section 6.5 Specifying a statistical model). The data are then used internally to evaluate the likelihood function for the model. The arguments are the name of the variable for which observations exist, and a list of data values:

 
Data(<identifier>, <value1>, <value2>, ...);

This specification can only be used with a matching Print() or PrintStep() for the same variable (see section 6.4.3 Print() specification; see section 6.4.4 PrintStep() specification). You must make sure that there are as many data values in the Data() specification as output time requested in the corresponding Print() or PrintStep(). A data value of "-1" is treated as "missing data" and ignored in likelihood calculations. The convention "-1" can be changed by changing INPUT_MISSING_VALUE in the header file `mc.h' and recompiling.


  webmaster     delorie software   privacy  
  Copyright 2003   by The Free Software Foundation     Updated Jun 2003