draw8sele: returns the candidate individuals for an empidata node.
In rebastaba: Handling Bayesian networks

Description Usage Arguments Details Value Examples

(dn) For a given individual (ind), the distances with all individuals in the rows of E are computed according to some requirement. From them a set of candidates of the E rows is returned. This set is comprised of candidates with distances greater or equal than di[1] and less or equal than di[2]. Their number is at least equal to nb[1] and no more than nb[2]. When the number of candidates is too small No candidates are returned; when it is too big the maximum number is returned with the smallest distances.
Details about the distance formula are given into the /rebastaba/ manual.

1	draw8sele(ind, E, wgt, kk, di, nb, nat)

`ind`	Value of the individual to predict (one row data.frame.
`E`	data.frame with the candidates in its rows.
`wgt`	the weights to be used for the distance computation.
`kk`	power coefficient for the distance computation.
`di`	extreme values of acceptable distances.
`nb`	minimum and maximum numbers of candidates.
`nat`	nature of the `ind` variables (named `character`).

No check are made

a data.frame with the same columns as those of 'E' containing the selected subset of candidates sorted according to the distance. Except for those with names given by rebastaba.ena[2] which are filled accordingly to the draw (their own distances).
When empty the returned data.frame has got no row.

 rebastaba3k("RESET"); # to comply R checking
 E <- data.frame(A=1:100, c=100:1); # the empirical distribution
 draw8sele(data.frame(A=20), E, structure(1, .Names="A"), 1, c(0, 1), c(1, 2), 
 structure("conti", .Names="A"));