PlotGLMERContinuous: Plot modelled effects of continuous variables with...

In timnewbold/StatisticalModels: What the package does (short line)

Plot modelled effects of continuous variables with uncertainty

Description

Plots estimated effects on a specified response variable of one or two specified continuous variables, from a generalized linear mixed-effects model. Where two continuous effects are specified, a perspective plot is used. Alternatively if byContEffect is used to specify the second effect, the first effect is shown at 3 levels of the second effect (2.5th percentile, median and 97.5th percentile).

Usage

PlotGLMERContinuous(model,data,effects,byContEffect=NULL,
                        otherContEffects=character(0),
                        otherFactors=character(0),xlab,ylab,
                        byFactor=NULL,zlab=NULL,outDir=NULL,
                        logLink="n",plotRug=FALSE,seMultiplier=1.96,
                        params=list(),add=FALSE,xlim=NULL,ylim=NULL,
                        zlim=NULL,line.cols=NULL,line.types=NULL,
                        plotUncertainty=TRUE,nPanels=1,main=NULL,
                        yDiv=1,transformX=FALSE)

Arguments

model	The model to derive parameter estimates from
data	A data frame containing all variables in the model
effects	The continuous variables (up to a maximum of 2) to plot an a continuous axis
byContEffect	An optional interacting continuous variable to plot at three levels (2.5th percentile, median and 97.5th percentile)
otherContEffects	Any continuous variables other than those selected for display that were fitted in the model. Median values of these variables will be used as the reference level.
otherFactors	Any factors that were fitted in the model, given as a list where the names are the names of the factors and the values are the factor levels for which to plot; shouldn't include any factors specified in byFactor
xlab	The x-axis label
ylab	The y-axis label
byFactor	Plot separate lines for each level in this factor
zlab	The z-axis label, for perspective plots showing two continuous effects
outDir	Optional: directory to write the graph to. If not specified, then the graph will be written to the native graphics device
logLink	The link function used by the model: one of "n" (identity), "e" (log, base e), "10" (log, base 10) or "b" (logit - used in binomial models)
plotRug	Whether to add a "rug" showing the values of the continuous variable represented in the data
seMultiplier	The number of standard errors to use for the error bars. Default is 1.96, which shows 95 percent confidence intervals
params	A list of R graphical parameters, which will overwrite the defaults
add	Whether to add the plot to an existing plot; default is FALSE
ylim	Fixed y-axis limits. If not specified, then resolved automatically
zlim	For plots with two continuous effects (perspective plots), optional limits for the z-axis
line.cols	The colours to use for the lines
line.types	The line types to use (solid, dashed, dotted etc.). Uses the same codes as R's graphical parameter 'lty'
plotUncertainty	Whether to include uncertainty estimates in the plot
nPanels	When byFactor is specified, the number of panels to divide the factor levels across
main	A title for the plot
yDiv	Factor to divide y-axis values by (useful for very large numbers, which would otherwise be printed in standard form)
transformX	Whether to back-transform log-transformed x-axis values to identity space. Currently only works with base-e

Details

Code for calculating predicted values and confidence intervals was taken from the GLMM wiki (see references).

Author(s)

Tim Newbold <[e-mail]>

References

http://glmm.wikidot.com/faq

Examples

# Load example data (site-level effects of land use on biodiversity from the PREDICTS database)
data(PREDICTSSiteData)

# Run a model of species richness as a function of land use, human population density
# and distance to nearest road (with an interaction between human population density
# and road distance)
m1 <- GLMER(modelData = PREDICTSSites,responseVar = "Species_richness",fitFamily = "poisson",fixedStruct = "LandUse+poly(logHPD.rs,2)+poly(logDistRd.rs,2)+poly(logHPD.rs,2):poly(logDistRd.rs,2)",randomStruct = "(1|SS)+(1|SSB)+(1|SSBS)",REML = TRUE)

# Plot the effect of human population density
PlotGLMERContinuous(model = m1$model,data = m1$data,effects = "logHPD.rs",otherContEffects = "logDistRd.rs",otherFactors = list(LandUse="Primary Vegetation"),xlab = "Human Population Density",ylab = "Species richness",logLink = "e",plotRug = TRUE)

# Plot a perspective plot of the effects on species richness of human population density
# and distance to nearest road
PlotGLMERContinuous(model = m1$model,data = m1$data,effects = c("logHPD.rs","logDistRd.rs"),otherFactors = list(LandUse="Primary Vegetation"),xlab = "Human population density",ylab = "Distance to road",zlab = "Species richness",logLink = "e",plotUncertainty = FALSE)

# Run a model of species as a function of land use, human population density, and their interaction
m1 <- GLMER(modelData = PREDICTSSites,responseVar = "Species_richness",fitFamily = "poisson",fixedStruct = "LandUse+poly(logHPD.rs,2)+LandUse:poly(logHPD.rs,2):",randomStruct = "(1|SS)+(1|SSB)+(1|SSBS)",REML = TRUE)

# Plot the effect on species richness of the interaction between human population density and land use
PlotGLMERContinuous(model = m1$model,data = m1$data,effects = "logHPD.rs",otherFactors = list(LandUse="Primary Vegetation"),xlab = "Human population density",ylab = "Species richness",byFactor = "LandUse",logLink = "e",line.cols=c("#E6AB02","#D95F02","#7570B3","#66A61E","#1B9E77","#E7298A"))
legend(0.75,15,c("Primary","Secondary","Plantation","Cropland","Pasture","Urban"),
col=c("#66A61E","#1B9E77","#7570B3","#E6AB02","#D95F02","#E7298A"),bty="n",lty=1)

timnewbold/StatisticalModels documentation built on Aug. 25, 2023, 4:58 p.m.