representative_point: Computes representative point based on the coordinates of...

In RADanalysis: Normalization and Study of Rank Abundance Distributions

Description

Computes representative point based on the coordinates of points which are in the same group.

Usage

representative_point(input, ids = NULL, coord_names = c(1, 2),
  standard_error_mean = TRUE, plot = FALSE, ...)

Arguments

input	A matrix which contains the coordinates of samples. Usually this is the result of ordination of normalized RADs using multi-dimensional scaling (cmdscale). In the input matrix each row contains vector of coordinates of one sample.
ids	Vector of row numbers of the desired group, from which a representative point is going to be represented
coord_names	A vector which contains the coordintes number that should be used to create representative point. Default is c(1,2).
standard_error_mean	A logical. If TRUE, uses the standard error of the mean and plot it with representative points. It works only if plot = TRUE.
plot	A logical. If TRUE, shows the representative points on the previous plot.
...	other graphical parameters to use for plotting. This function uses internally the functions points and arrows to plot.

Value

A list of following parameters:

$mean: Contains the average of points. A vector with the length of coordinates used for computing the average. These coordinates are preset in coord_names.

$sd: A vector with a length similar to mean which contains the standard deviation for each coordinate.

$mean_standard_error: A vector with a length similar to mean which contain the standard deviation of the mean for each coordinate. This vector is the result of sd / sqrt(n), when n is the number of members of the group (length of sample_ids).

If plot = TRUE, representative points would be added to the previous plot.

If standard_error_mean = TRUE, the standard error of the mean would be added to the representative points.

See Also

RADnormalization for normalize an abundance vector. This function return more details compared to RADnormalization_matrix, RADnormalization_matrix for normalize an entire otutable, representative_RAD for study the representative of group of norm rads.

Examples

line_cols <- c("green","red","blue")
sample_classes <- c(1,1,1,1,2,2,3,3,1,1,2,3,3,1,1,2,3,3)
maxrank <- 400
data("gut_nrads")
nrads <- gut_nrads
nrads <- nrads$norm_matrix

#distance matrix using manhattan distance
d <- dist(x = nrads,method = "manhattan")
#ordination using classical multi-dimensional scaling
mds <- cmdscale(d = d,k = 5,eig = TRUE)

#plot the points
plot(mds$points,xlab = "First coordinate",ylab = "Second coordinate",pch = 19,cex =1,
    col = line_cols[sample_classes],
    main = "MDS plot with representative points \n of each group and error bars")

#add the representative points wit erorr bar to the previous plot
a <- representative_point(input = mds$points,ids = which(sample_classes == 1),
    col = scales::alpha(line_cols[1],0.5),
    plot = TRUE,standard_error_mean = TRUE,pch = 19, cex = 4)
a <- representative_point(input = mds$points,ids = which(sample_classes == 2),
    col = scales::alpha(line_cols[2],0.5),
    plot = TRUE,standard_error_mean = TRUE,pch = 19, cex = 4)
a <- representative_point(input = mds$points,ids = which(sample_classes == 3),
    col = scales::alpha(line_cols[3],0.5),
    plot = TRUE,standard_error_mean = TRUE,pch = 19, cex = 4)

legend("bottomleft",bty = "n",legend = c("pre Cp","under Cp","post Cp"),
    col = line_cols,pch = 19)

RADanalysis documentation built on May 2, 2019, 6:13 a.m.