Compare matrices via Selection Response Decomposition
Based on Random Skewers tecnique, selection response vectors are expanded in direct and indirect componenet by trait and compared via vector correlations.
SRD(cov.x, cov.y, ...) ## Default S3 method: SRD(cov.x, cov.y, iterations = 1000, ...) ## S3 method for class 'list' SRD(cov.x, cov.y = NULL, iterations = 1000, parallel = FALSE, ...) ## S3 method for class 'SRD' plot(x, matrix.label = "", ...)
cov.x |
Covariance matrix being compared. cov.x can be a matrix or a list. |
cov.y |
Covariance matrix being compared. Ignored if cov.x is a list. |
... |
aditional parameters passed to other methods |
iterations |
Number of random vectors used in comparison |
parallel |
if TRUE computations are done in parallel. Some foreach backend must be registered, like doParallel or doMC. |
x |
Output from SRD function, used in ploting |
matrix.label |
Plot label |
Output can be ploted using PlotSRD function
List of SRD scores means, confidence intervals, standard deviations, centered means e centered standard deviations
pc1 scored along the pc1 of the mean/SD correlation matrix
model List of linear model results from mean/SD correlation. Quantiles, interval and divergent traits
If input is a list, output is a symmetric list array with pairwise comparisons.
Diogo Melo, Guilherme Garcia
Marroig, G., Melo, D., Porto, A., Sebastiao, H., and Garcia, G. (2011). Selection Response Decomposition (SRD): A New Tool for Dissecting Differences and Similarities Between Matrices. Evolutionary Biology, 38(2), 225-241. doi:10.1007/s11692-010-9107-2
cov.matrix.1 <- cov(matrix(rnorm(30*10), 30, 10)) cov.matrix.2 <- cov(matrix(rnorm(30*10), 30, 10)) colnames(cov.matrix.1) <- colnames(cov.matrix.2) <- sample(letters, 10) rownames(cov.matrix.1) <- rownames(cov.matrix.2) <- colnames(cov.matrix.1) srd.output <- SRD(cov.matrix.1, cov.matrix.2) #lists m.list <- RandomMatrix(10, 4) srd.array.result = SRD(m.list) #divergent traits colnames(cov.matrix.1)[as.logical(srd.output$model$code)] #Plot plot(srd.output) ## For the array generated by SRD(m.list) you must index the idividual positions for plotting: plot(srd.array.result[1,2][[1]]) plot(srd.array.result[3,4][[1]]) #Multiple threads can be used with some foreach backend library, like doMC or doParallel #library(doParallel) ##Windows: #cl <- makeCluster(2) #registerDoParallel(cl) ##Mac and Linux: #registerDoParallel(cores = 2) #SRD(m.list, parallel = TRUE)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.