Description Usage Arguments Details Value Examples
Calculation of accuracy for specified thresholds.
evaluate
is called to perform the calculations.
1 2 3 
x 
an object of class 
te.u 
the data to predict on. 
te.y 
a vector of observed positive/negative (presence/absence) values 
positive 
the positive label in te.y. if this is not given, the class with the lower frequency is assumed to be the positive class. 
th 
a vector of thresholds at which the model is evaluated 
allModels 
logical, default is 
modParam 
data frame with the parameters of the model to be evaluated 
modRow 
the row of the model in the 
modRank 
the rank of the model after sorting by 
by 
character. must be a metric available in the 
decreasing 
only when 
verbose 

... 
arguments passed to 
By defalut, only the final model is evaluated when allModels
is FALSE
and non of the arguments modParam
,modRow
, or modRank
given.
an object of class ModelEvaluation
(see ModelEvaluationclass
))
or ModelSelectionEvaluation. The latter is a list with two elements, the
first containing the model sleetion table and the second a list with the evaluation
results, each of whith a ModelEvaluation
object.
The rows in model selection table correspond to the evaluation list elements.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30  ## Not run:
# get training and test data
data(bananas)
seed < 123456
tr.x < bananas$tr[, 1]
tr.y < bananas$tr[, 1]
set.seed (seed)
te.i < sample ( ncell (bananas$y), 1000 )
te.x < extract (bananas$x, te.i)
te.y < extract (bananas$y, te.i)
# run trainOcc
oc < trainOcc(x=tr.x, y=puFactor(tr.y),
tuneGrid=expand.grid(sigma=c(0.1,1), ### not so large grid
cNeg=2^seq(5, 10, 3),
cMultiplier=2^seq(4, 15, 2)))
# evaluate the final model
ev < evaluateOcc(oc, y=te.y, te.u=te.x)
# besides the thresholds used, this is identical to:
te.pred < predict(oc, te.x)
ev < evaluate(p=te.pred[te.y==1], a=te.pred[te.y!=1])
# evaluate several models
# e.g. evaluate models with a true positive rate (tpr) higher than 0.8 and a
# positive prediction probability (ppp) small than 0.4
modRows < which(oc$results$tpr>=0.8 & oc$results$ppp<0.4)
ev < evaluateOcc(oc, y=te.y, te.u=te.x, modRow=modRows)
# plot the puperformance metric versus the maximum kappa
evList < print(ev)
plot(evList$puF, evList$mxK.K, xlab="puF", ylab="max. Kappa")
## End(Not run)

Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.