R/distdisplayed.R

In BiodiversityR: Package for Community Ecology and Suitability Analysis

`distdisplayed` <-
function(x, ordiplot, distx="bray", plotit=T, addit=F, method="spearman", permutations=100, abline=F, gam=T,...) {
    if (gam==T) {
#        if (!require(mgcv)) {stop("Requires package mgcv")}
    }
    if (inherits(x, "dist")) {
        dist1 <- x
        xlab <- attr(x,"method")
        if (is.null(xlab)) {
            xlab <- "distance matrix 1"
        }else{
            xlab <- paste(xlab, "distance")
        }
    }else{
        dist1 <- vegdist(x, method = distx)
        xlab <- "original distance"
    }
    if (inherits(ordiplot, "dist")) {
        dist2 <- ordiplot
        ylab <- attr(dist2,"method")
        if (is.null(ylab)) {
            ylab <- "distance matrix 2"
        }else{
            ylab <- paste(ylab, "distance")
        }
    }else{
        ordiscores <- scores(ordiplot,display="sites")
        dist2 <- vegdist(ordiscores,method="euclidean")
        ylab <- "distance in ordination plot"
    }
    if (plotit==T) {
        if (addit==F) {
            graphics::plot(dist1, dist2, xlab=xlab, ylab=ylab)
        }
        if (abline==T) {abline(0,1)}
        if (gam==T){
            data <- data.frame(cbind(as.numeric(dist1), as.numeric(dist2)))
            names(data) <- c("dist1", "dist2")
            seq <- order(data[,1])
            sorted <- data
            sorted[1:nrow(data),] <- data[seq,]
            gamresult <- mgcv::gam(as.formula(dist2 ~ s(dist1)), data=sorted)
            newdata <- data.frame(seq(min(sorted[,1]), max(sorted[,1]), length = 1000))
            colnames(newdata) <- "dist1"
            gamresult2 <- predict(gamresult, newdata) 
            graphics::points(newdata$dist1,gamresult2,type="l",lwd=2,col="red")
        }
    }
    result2 <- mantel(dist1, dist2, method=method, permutations=permutations,...)
    if (gam==T) {
        return(list(gamanalysis=summary(gamresult), mantelanalysis=result2))
    }else{
        return(result2)
    }
}