cor.matrix: Function to obtain the correlation between two matrices and partial matrix correlation between three matrices.

Description

The functions cor.matrix and cor.matrix.partial are similar the function mantel and mantel.partial, although the significance of the statistics is evaluated differently from Mantel. The functions pro.matrix and pro.matrix.partial use symmetric Procrustes as a measure of concordance between data sets. The function cor.mantel is similar to the function mantel, but allows the use of a set of predefined permutation. For more details, see syncsa.

Usage

cor.matrix(
  mx1,
  mx2,
  x,
  my1 = NULL,
  my2 = NULL,
  y,
  permute.my2 = FALSE,
  method = "pearson",
  dist = "euclidean",
  permutations = 999,
  norm = FALSE,
  norm.y = FALSE,
  strata = NULL,
  na.rm = FALSE,
  seqpermutation = NULL,
  parallel = NULL,
  newClusters = TRUE,
  CL = NULL
)
cor.mantel(
  dist.x,
  dist.y,
  method = "pearson",
  permutations = 999,
  strata = NULL,
  na.rm = FALSE,
  seqpermutation = NULL,
  parallel = NULL,
  newClusters = TRUE,
  CL = NULL
)
cor.matrix.partial(
  mx1,
  mx2,
  x,
  my1 = NULL,
  my2 = NULL,
  y,
  mz1 = NULL,
  mz2 = NULL,
  z,
  method = "pearson",
  dist = "euclidean",
  permute.my2 = FALSE,
  permute.mz2 = FALSE,
  permutations = 999,
  norm = FALSE,
  norm.y = FALSE,
  norm.z = FALSE,
  strata = NULL,
  na.rm = FALSE,
  seqpermutation = NULL,
  parallel = NULL,
  newClusters = TRUE,
  CL = NULL
)
cor.matrix2(
  mx1,
  mx2,
  x,
  y,
  method = "pearson",
  dist = "euclidean",
  put.together = NULL,
  permutations = 999,
  strata = NULL,
  na.rm = FALSE,
  seqpermutation = NULL,
  parallel = NULL,
  newClusters = TRUE,
  CL = NULL
)
cor.procrustes(
  x,
  y,
  permutations = 999,
  strata = NULL,
  na.rm = FALSE,
  seqpermutation = NULL,
  parallel = NULL,
  newClusters = TRUE,
  CL = NULL
)
pro.matrix(
  mx1,
  mx2,
  x,
  my1 = NULL,
  my2 = NULL,
  y,
  permute.my2 = FALSE,
  permutations = 999,
  norm = FALSE,
  norm.y = FALSE,
  strata = NULL,
  seqpermutation = NULL,
  parallel = NULL,
  newClusters = TRUE,
  CL = NULL
)
pro.matrix.partial(
  mx1,
  mx2,
  x,
  my1 = NULL,
  my2 = NULL,
  y,
  mz1 = NULL,
  mz2 = NULL,
  z,
  permute.my2 = FALSE,
  permute.mz2 = FALSE,
  permutations = 999,
  norm = FALSE,
  norm.y = FALSE,
  norm.z = FALSE,
  strata = NULL,
  seqpermutation = NULL,
  parallel = NULL,
  newClusters = TRUE,
  CL = NULL
)
pro.matrix2(
  mx1,
  mx2,
  x,
  y,
  put.together = NULL,
  permutations = 999,
  strata = NULL,
  seqpermutation = NULL,
  parallel = NULL,
  newClusters = TRUE,
  CL = NULL
)

Arguments

mx1

Matrix that multiplied by mx2 results in the matrix x.

mx2

Matrix that when multiplied by mx1 results in the matrix x. See `details` below.

Matrix that will be correlated with the matrix y.

my1

Matrix that multiplied by my2 results in the matrix y.

my2

Matrix that when multiplied by my1 results in the matrix y. See `details` below.

Matrix that will be correlated with the matrix x.

permute.my2

Logical argument (TRUE or FALSE) to specify if realize parallel permutation in matrix my2.

method

Correlation method, as accepted by cor: "pearson", "spearman" or "kendall".

dist

Dissimilarity index, as accepted by vegdist: "manhattan", "euclidean", "canberra", "bray", "kulczynski", "jaccard", "gower", "altGower", "morisita", "horn", "mountford", "raup" , "binomial" or "chao".

permutations

Number of permutations in assessing significance.

norm

Logical argument (TRUE or FALSE) to specify if x is standardized within variables (Default norm = FALSE).

norm.y

Logical argument (TRUE or FALSE) to specify if y is standardized within variables (Default norm = FALSE).

strata

Argument to specify restricting permutations within species groups (Default strata = NULL).

na.rm

Logical argument (TRUE or FALSE) to specify if pairwise deletion of missing observations when computing dissimilarities (Default na.rm = FALSE).

seqpermutation

A set of predefined permutation, with the same dimensions of permutations (Default seqpermutation = NULL).

parallel

Number of parallel processes. Tip: use parallel::detectCores() (Default parallel = NULL).

newClusters

Logical argument (TRUE or FALSE) to specify if make new parallel processes or use predefined socket cluster. Only if parallel is different of NULL (Default newClusters = TRUE).

A predefined socket cluster done with parallel package.

dist.x

Dissimilarity matrices of class dist.

dist.y

Dissimilarity matrices of class dist.

mz1

Matrix that multiplied by mz2 results in the matrix z.

mz2

Matrix that when multiplied by mz1 results in the matrix z. See `details` below.

Matrix whose effect will be removed from the correlation between x and y.

permute.mz2

Logical argument (TRUE or FALSE) to specify if realize parallel permutation in matrix mz2.

norm.z

Logical argument (TRUE or FALSE) to specify if z is standardized within variables (Default norm = FALSE).

put.together

List to specify group of traits. Each group specify receive the same weight that one trait outside any group, in the way each group is considered as unique trait (Default put.together = NULL). This argument must be a list, see examples in syncsa.

Value

Obs

Correlation between matrices.

Significance level based on permutations.

Details

The null model is based on permutations in the matrix mx2, typically the matrices B, U and Q, except in the function cor.mantel when the permutations are done in one of distance matrix.

Null model described by Pillar et al. (2009) and Pillar & Duarte (2010). For more details on the matrices and the null model, see syncsa.

References

Pillar, V.D.; Duarte, L.d.S. (2010). A framework for metacommunity analysis of phylogenetic structure. Ecology Letters, 13, 587-596.

Pillar, V.D., Duarte, L.d.S., Sosinski, E.E. & Joner, F. (2009). Discriminating trait-convergence and trait-divergence assembly patterns in ecological community gradients. Journal of Vegetation Science, 20, 334:348.