ReliabilityPlot: Produces are relaibility plot with associate linear...

Description Usage Arguments Details Data type Version Date submitted Author(s) See Also

View source: R/ReliabilityPlot.R

Description

Returns a reliability plot of the discritised predictions versus obserations as well as the a vector of the intercept, gradient, and r-squared for further interpretation

Usage

1
ReliabilityPlot(.model, .ras, n_bins = 20)

Arguments

.model

Internal parameter, do not use in the workflow function. .model is list of a data frame (data) and a model object (model). .model is passed automatically in workflow, combining data from the model module(s) and process module(s), to the output module(s) and should not be passed by the user.

.ras

Internal parameter, do not use in the workflow function. .ras is a raster layer, brick or stack object. .ras is passed automatically in workflow from the covariate module(s) to the output module(s) and should not be passed by the user.

n_bins

An integer greater than 1 indicating the number of bins for predictions and obsertions

Details

We can analyse the linear model of predictions versus observations for bias in our predictions. A gradient on 1 and and intercept of 0 indicate a perfect fit - the predictions equal the observations. An intercept less than 0 indicated that, on average, our model underpredicts. Greater than 0 indicates over-prediction, on average.

Data type

presence-only, presence/absence, presence/background

Version

0.1

Date submitted

2018-02-27

Author(s)

Liz Martin, lizmartinresearch@gmail.com

See Also

Other output: AIC, AUC, Appify, CoefficientPlot, CovHistograms, DataSummary, Deviance, GenerateCovariateReport, InteractiveCovariateMap, InteractiveMap, InteractiveOccurrenceMap, NoOutput, PairPlot, PerformanceMeasures, PredictNewRasterMap, PrintMap, PrintOccurrenceMap, ROCcurve, ResponseCurveViz, ResponseCurve, ResponsePlot, SameTimePlaceMap, SeparatePA, SurfaceMap, VariableImportance


zoonproject/modules documentation built on May 4, 2019, 11:25 p.m.