Each module uses a common parameter file to define options. This parameter file should reside in the top-level directory (one level up from each module directory). Please refer to individual module pages for a more detailed description on what parameters, if any, need to be set. You can download a sample parameters file here. Below is an example of a parameter file, along with commented descriptions of how each parameter is used:
============================ Begin varFile.var Example ==================================== #-- file names --# trainingFile = asia_train.xml # A file of training examples (network instantiations) # used with K2 and Glue valFile = asia_val # A file of validating network instantiations used # with Glue printingsFile = printings.csv # The specified output file for RMSE values from Glue # and StochasticSampling. If left blank, # output will be directed to printings.txt outputNetworkFile = output.xml # The learned network from K2 goldNetwork = asia.xml # The "gold standard" network used with K2 and Glue evidence_filename = evidence.txt # The evidence filename used with Glue, # StochasticSampling and LS cOrder = 0 2 5 6 3 1 4 7 # The initial ordering of the network for K2 and Glue # a b c d e f h g #-- glue params --# port=3200 # The port number for running Glue populationSize=100 # The population size for the genetic algorithm in Glue generations = 100 # The number of generations for the genetic algorithm in Glue outputFile = output #-- sampling params --# sampling_method = -l # The type of sampling in StochasticSampling network_stem = asia # The "stem" of the Bayesian network for StochasticSampling new_or_old = old # When using Glue, this should be "new". Otherwise, set to "old" casenumber = 5000 # How many samples to take for K2, StochasticSampling, and Glue ============================ End varFile.var Example =======================================