new8alk: generates a new alk from user specification
In rebastaba: Handling Bayesian networks

Description Usage Arguments Details Value

(ba) Just collecting specifications provided by the user to build an alk object up taking into consideration the specificity of the ltype.
Default values are consistent for not to be used arguments.

new8alk(ldes, ltype, lpara=rbsb.lis0, lrep=0, lnat=rbsb.cha0,
	  lvar=rbsb.cha0, lparent=rbsb.cha0, lpod=rbsb.lis0,
	  lred=rbsb.lis0, lcod=rbsb.lis0, ltransfo="",
	  ldaf=rbsb.daf0, lwin=rebastaba.win0, lfunct=rbsb.fun0)

`ldes`	Human description of the potential node through either a character for the name or a /des/ object.
`ltype`	type of link: a character string which must belong to the `names(rebastaba.tlk)`. This list is returned by `help8ltype("ls")`.
`lpara`	list of parameter values for standard distributions.
`lrep`	the number of repetitions; zero means no repetitions. Not all node types can be repeated.
`lnat`	According to the ltype (see `help8ltype`) can be any component of `rbsb.sna`. When rbsb.cha0, the standard nature associated to `ltype`. When it is a multivariate node, it must be a vector with the corresponding natures. For 'empidata' it is given by the slot `@nat` of `lwin`.
`lvar`	This argument can be used either for multivariate nodes, or for univariate nodes which are repeated (indeed it is the way to indicate repetitions which are then possible only for univariate distribution. In both case, lvar provides the vector of variable names of the node. Its length gives the number of variables within the added node. Standard probability distributions are associated with a number of variables and it will be checked. For 'empidata' it is given by the names of slot `@nat` of `lwin`.
`lparent`	parent vector at the node level; in most cases these are deduced from the other arguments.
`lpod`	When rbsb.lis0, possible domain(s) associated to ltype is provided (see rebastaba.tlk). This must be a list with as many components as variables in the added node. Each component of the list is associated with a variable of the node (in the same order as given in slot lvar. When there is only one variable, can be a vector instead of a list with a unique element which is a vector. () when lnat== 'cateo' or 'categ', as many character strings as the number of categories. Be aware that the length of lpod provides also the number of categories! () when lnat=="integ" the lower and upper bounds of the variation of the integer variable (e.g. c(0,17)) (*) when lnat=="conti" the strict lower and upper bounds of the variation of the continuous variable (e.g. c(-I
`lred`	Idem of lpod but for representation domain instead of possible domaine
`lcod`	Idem of lpod but for common domain instead of possible domaine
`ltransfo`	a simple way to define a transformation of the drawn values by some easyprogramming string. Also "[\|2\|]" means rounding with 2 decimals and remind that the alias for the node under creation is "Y".
`ldaf`	in the case of data.frame distributions provides the associated /daf/.
`lwin`	/win/ object describing the way to use a data frame based distribution. See the description of this object for more details.
`lfunct`	can be a function to be used as simulating function when ltype is `program`.