plot8dn: Plots the numeric or categoric variables of a dn object
In rebastaba: Handling Bayesian networks

Description Usage Arguments Details Value Examples

(dn) produces the (X,Y) plots from a given dn with possible selection of variables and simulations. red and cod slots are used. To get a better picture for cartesian representations, some shaking can be applied to the integ variables.
When X!=Y in the couple (X,Y) a bivariate representation is issued according to the nature of the two variables. When X==Y, then a univariate representation is performed.

1
2
3

plot8dn(x, y="useless", ..., titles=NULL, stitles=NULL, xxlab=NULL,
	  yylab=NULL, varx=1, vary=0, quels=0, lab=1, shaking=0.100,
	  file="")

`x`	the rbsb object to plot
`y`	For compatibility with the generic plot function. Useless here.
`...`	Further arguments to be passed to the plot function.
`titles`	Defines the titles to provide on each Cartesian diagram(s) with a character vector cycled as necessary to cover all diagrams. If NULL the name of the dn with pertinent information is issued.
`stitles`	Defines the subtitles to provide on each Cartesian diagram(s) with a character. If NULL standard sub-title.
`xxlab`	As `stitles` but for `xlab` arguments of the plots.
`yylab`	As `xxlab` for `ylab`.
`varx`	Indicates the variables to use as abscissae. If 1, all numeric variables of the dn are involved. If 2, all categoric variables of the dn are involved. If 3, all variables of the dn are involved. If not, indicates by their names which variables must be used.
`vary`	Indicates which variables to use as ordinates. If -1, only univariate graphs of `varx` are done, If 0, `vary` is set identical to `varx`, If 1, all numeric variables of the dn are involved, If 2, all categoric variables of the dn are involved, If 3, all variables of the dn are involved. For these numerical orders, reciprocal bivariate graphs are discarded. If not numeric, must have the same size as varx providing the names of the variables to use in ordinates.
`quels`	Indicates the selection of the points to be displayed. (i) 0 then all simulations are displayed. (ii) A positive integer indicating the number of simulations to sample within the data frame. (iii) Can be a character string and must be on the format "V1==3" indicating that only the simulated observations for which the variable "V1" has got 3 as value will be displayed. So "V3" must exist and probably is an integ or categ variable. Be careful that no spaces are possible, e.g. "V1 == 3" will not work. (iv) Can be a boolean vector with as many components as simulated observations, only those with TRUE will be selected. (v) When its length is zero, no plot is performed.
`lab`	Indicates the labels with which the points must be identified in case of binumeric plots. (i) Can be a numeric scalar: the symbol for localization. (ii) Can be a numerical vector with the same size that the number of rows of x@df, used to indicate which standard symbols of R plot to use as label (in case of selection, only those selected will be used). (iii) Can be a single character: then must indicates a name variables (if continuous the values of it will be used after rounding as labels). (iv) Can be a character vector with the same size that the number of rows of x@df, used as there are (in case of selection, only those selected will be used).
`shaking`	In standard deviation of each variable, the coefficient for giggling the 'integ' variables in order to better see identically plotted points in binumeric plots.
`file`	When not empty produces a text file of this name containing the numbering of all page representations. Quite useful to find a specific diagram when the diagrams are numerous.

Some

returns nothing but plots are drawn and possibly a text file is created.

 rebastaba3k("RESET"); # for R checking convenience
 plot(rebastaba.dn1);
 plot(rebastaba.dn1, vary=-1);
 plot(rebastaba.dn4, varx="A", vary="B");
 plot(rebastaba.dn4, varx="A", vary="D");
 plot(rebastaba.dn4, varx="D", vary="A");
 plot(rebastaba.dn4, varx="E", vary="F");