create

Overview

The create command is used to create objects and assign them to aliases. The alias must not already be in use.

Applies To

All object types.

Syntax

general syntax

create ObjectType alias NewAlias Arguments

where:
ObjectType is the type of object to create
alias is the alias to assign the new object to

BPTrainer

create BPTrainer alias NewAlias network NetworkAlias

where:
network is the alias of the MLP or FuNN to train

Data

create Data alias NewAlias inputs #Inputs outputs #Outputs rows #Rows labels on / off

where:
inputs is the number of columns in the input table
outputs is the number of columns in theoutput table
rows is the number of storage rows to allocate
labels indicates whether of not labels are present

DataStats

create DataStats alias NewAlias data DataAlias

where:
data is the alias of the Data object of which to calculate the statistics

FuNN

create FuNN alias NewAlias inputs #Inputs condition {InputMFArray} rules #RuleNodes action {OutputMFArray} outputs #Outputs

where:
inputs is the number of input nodes
condition is an array specifying the number of condition nodes attached to each input
rules is the number of rule nodes
action is an array specifying the number of action nodes attached to each output
outputs is the number of output nodes

GATrainer

create GATrainer alias NewAlias network NetworkAlias minweight MinimumWeight maxweight MaximumWeight population Popsize

where:
network is the alias of the MLP or FuNN to train
minweight is the minimum allowable value of connection weights
maxweight is the maximum allowable value of connection weights
population is the number of individuals in the training population

Kohonen

create Kohonen alias NewAlias dimensions #MapDimensions inputs #Inputs mapdim {MapDimensionsArray}

where:
dimensions is the dimensionality of the output map
inputs is the umber of input nodes
mapdim is an array specifying the size of each dimension of the output map

KohonenTrainer

create KohonenTrainer alias NewAlias network NetworkAlias

where:
network is the alias of the Kohonen to train

LWFTrainer

create LWFTrainer alias NewAlias network NetworkAlias

where:
network is the alias of the FuNN to train by LWF

MLP

create MLP alias NewAlias layers #Layers inputs #Inputs hidden {HiddenLayersSizes} outputs #Outputs bias on / off

where:
layers is the number of layers of nodes in the MLP
inputs is the number of input nodes in the MLP
hidden is an array of values describing the size of each hidden layer
outputs is the number of output nodes
bias indicates whether or not the MLP has a bias

This page is maintained by Michael Watts (http://mike.watts.net.nz)
Last modified on: 12/10/98.