library(frbs)

## Load the Iris data
data(iris)

## Shuffle the data
irisShuffled <- iris[sample(nrow(iris)), ]

## Convert the response variable into numeric
irisShuffled[,5] <- unclass(
   irisShuffled[, 5])
   
## Split the data into training and testing datasets
tra.iris <- irisShuffled[1 : 105, ]
tst.iris <- irisShuffled[106 : 
   nrow(irisShuffled), 1:4]
   
## Save actual values
real.iris <- matrix(irisShuffled
  [106 : nrow(irisShuffled), 5], ncol = 1)
  
## Define the range of input variables
range.data.input <- apply(iris[,-ncol(iris)],
  2, range)

## Set parameter values
method.type <- "FRBCS.W"
control <- list(num.labels = 3, 
  type.mf ="TRAPEZOID", type.tnorm = "MIN", 
  type.snorm = "MAX", 
  type.implication.func = "ZADEH")

## Construct an FRBS model
mod.class <- frbs.learn(tra.iris, 
  range.data.input, method.type, control)

## We can just display the frbsPMML format
modPMML <- frbsPMML(mod.class)
  
## Export and save into "modClass.frbsPMML"
write.frbsPMML(modPMML, "modClass")

## Import the frbsPMML format back to the FRBS model
objectClass <- read.frbsPMML("modClass.frbsPMML")

## Predict new data (tst.iris)
res.test <- predict(objectClass, tst.iris)

## Calculate error
err = 100 * sum(real.iris != res.test)/
  nrow(real.iris)
print(err)