R Graphical Manual

Browse All

Last data update: 2014.03.03

R: From a set of rules, remove each rule for which another rule...

perceive

R Documentation

From a set of rules, remove each rule for which another rule exists that is more specific.

Description

Examine rules in a list and remove all of them for whose other more specific rules are present in the list. The specificity is determined by calling the is.specific function. This operation is a part of the pbld inference mechanism.

Usage

perceive(rules, 
         vars,
         specs,
         type=c('global', 'local'),
         fired=NULL)

Arguments

`rules`	A list of character vectors where each element is a fuzzy set name (a predicate) and thus each such vector forms a rule.
`vars`	A named (typically character) vector that determines which predicates originate from the same variable, i.e. which of them semantically deal with the same property. For that purpose, each value from any vector stored in the `rules` list must be present in `names(vars)`. See also `vars` function of the `fsets` class.
`specs`	A square numeric matrix containing values from {0, 1}. It is a specificity matrix for which each row and column corresponds to an `rules`'es predicate `specs[i][j] = 1` if and only if the i-th predicate is more specific (i.e. the corresponding fuzzy set is a subset of) than the j-th predicate (i.e. `x[, j]`). See also `specs` function of the `fsets` class.
`type`	The type of perception to use. It can be either `"local"` or `"global"` (default).
`fired`	If `type=="global"` then this argument can be NULL. If `type` is `"local"` then `fired` must be a numeric vector of values in the interval [0,1] indicating the truth values of all rules, i.e. the length of the vector must be equal to the number of rules in the `rules` argument.

Details

For each rule x in the rules list, it searches for another rule y such that is.specific(y, x) returns TRUE. If yes then x is removed from the list.

Value

A modified list of rules for which no other more specific rule exists. (Each rule is a vector.)

Author(s)

Michal Burda

Examples

# prepare vars
v <- c(rep('a', 3),
       rep('b', 3),
       rep('c', 3),
       rep('d', 3))
names(v) <- paste(rep(c('VeSm', 'Sm', 'Bi'), times=4),
                  rep(c('a', 'b', 'c', 'd'), each=3),
                  sep='.')
print(v)

# prepare specs
s <- matrix(c(0,1,0, 0,0,0, 0,0,0, 0,0,0,
              0,0,0, 0,0,0, 0,0,0, 0,0,0,
              0,0,0, 0,0,0, 0,0,0, 0,0,0,

              0,0,0, 0,1,0, 0,0,0, 0,0,0,
              0,0,0, 0,0,0, 0,0,0, 0,0,0,
              0,0,0, 0,0,0, 0,0,0, 0,0,0,

              0,0,0, 0,0,0, 0,1,0, 0,0,0,
              0,0,0, 0,0,0, 0,0,0, 0,0,0,
              0,0,0, 0,0,0, 0,0,0, 0,0,0,

              0,0,0, 0,0,0, 0,0,0, 0,1,0,
              0,0,0, 0,0,0, 0,0,0, 0,0,0,
              0,0,0, 0,0,0, 0,0,0, 0,0,0),
          byrow=TRUE,
          ncol=12)
colnames(s) = names(v)
rownames(s) = names(v)
print(s)

# run perceive function: (Sm.a, Bi.c) has
# more specific rule (VeSm.a, Bi.c)
perceive(list(c('Sm.a', 'Bi.c'), 
              c('VeSm.a', 'Bi.c'),
              c('Sm.b', 'Sm.d')),
         v, s)