enmtools.calibrate {ENMTools} | R Documentation |
Takes an emtools.model object, and reformats it to run through the CalibratR package, calculates Continuous Boyce Index, and runs a Hosmer-Lemeshow goodness-of-fit test. Can either do a full CalibratR run or just return ECE/MCE statistics and plots.
Description
Takes an emtools.model object, and reformats it to run through the CalibratR package, calculates Continuous Boyce Index, and runs a Hosmer-Lemeshow goodness-of-fit test. Can either do a full CalibratR run or just return ECE/MCE statistics and plots.
Usage
enmtools.calibrate(
model,
recalibrate = FALSE,
cuts = 11,
env = NA,
n.background = 10000,
recal.methods = c(1, 2, 3, 4),
...
)
Arguments
model |
An enmtools.model object |
recalibrate |
When TRUE, does a full CalibratR "calibrate" run to recalibrate the model. When FALSE, just returns metrics and plots measuring calibration of the model as is. |
cuts |
The number of bins to split suitability scores into for calculating calibration. |
env |
A set of environment layers to be used for optional env space metrics |
n.background |
Number of background points to be used for env space metrics |
recal.methods |
Numeric or vector specifying which methods the CalibratR package should use for recalibration. 1=hist_scaled, 2=hist_transformed, 3=BBQ_scaled, 4=BBQ_transformed. Option 5 is disabled by default due to errors. |
... |
Further arguments to be passed to CalibratR's "calibrate" function. |
Value
An enmtools.calibrate object containing calibration and classificaction plots, metrics of model calibration, and (optionally) versions of the model that were recalibrated using various methods.
Examples
#install.extras(repos='http://cran.us.r-project.org')
monticola.glm <- enmtools.glm(iberolacerta.clade$species$monticola,
env = euro.worldclim,
f = pres ~ bio1 + bio9,
test.prop = 0.3)
if(check.extras("enmtools.calibrate")) {
enmtools.calibrate(monticola.glm)
}