predict.FPM {RFPM}

Predict Toxicity Using the Floating Percentile Model

Description

Use new sediment chemistry data to generate FPM predictions

Usage

## S3 method for class 'FPM'
predict(object, newdata, ...)

Arguments

object

FPM class object, created using FPM (and extracted from the resulting list using .$FPM, where '.' is the name of the object or FPM call)

newdata

character vector of column names of chemical concentration variables in data

...

further arguments passed to or from other methods

Details

There are two things to keep in mind when using predict for 'FPM' objects. Firstly, when FPM is run, the output object is a list; one of the objects (called "FPM") is the FPM class object that can be used to predict Hits.

Secondly, unlike other default "predict" methods, predict.FPM is used strictly to predict toxicity that is not included in the original dataset used to generate fpm. To predict Hit results for the original data, set hitInfo == TRUE when running the FPM function. Note that predicting toxicity for the original dataset may be arbitrary/unnecessary, as the Hit results for the original dataset must be known.

Note that, in order to run predict.FPM, the newdata argument must be supplied a data.frame that includes at least one sample with all of the chemical columns contained in fpm. Column headers must match exactly.

Value

logical

Examples

# create FPM object with chemical headings that overlap
overlap <- intersect(names(h.northport)[1:10], 
            names(h.tristate)[7:13])
fpm_object = FPM(h.northport, overlap)
# run predict on the 'FPM' list item to estimate Hits
predict(object = fpm_object$FPM, newdata = h.tristate)