corMatCCA: Perform Canonical Correlation Analysis on a correlation...

In parhammarstrom/multivariateAnalysisVT2021: Package for excercises in Multivariate Analysis.

Description

This function performs a Canonical Correlation Analysis on a correlation matrix. It assumes that the correlation matrix contains variable column names. If no 'nObs' is supplied, no significance test will be performed.

Usage

corMatCCA(
  corMat,
  p,
  set1name = "set1",
  set2name = "set2",
  useColumnNames = T,
  set1VarPrefix = "U",
  set2VarPrefix = "V",
  nObs = NA,
  alpha = 0.05
)

Arguments

corMat	A symmetric correlation matrix.
p	The number of variables in the first set. Used to separate the sets.
set1name	Name of the first set. Defaults to "set1".
set2name	Name of the second set. Defaults to "set2".
useColumnNames	Boolean that indicates whether to use the column names in the supplied correlation matrix. Defaults to True.
set1VarPrefix	Prefix used for the variables in set 1. Ignored if useColumnNames=True and if all columns in the correlation matrix contain variable names. Defaults to "U".
set2VarPrefix	Prefix used for the variables in set 2. Ignored if useColumnNames=True and if all columns in the correlation matrix contain variable names. Defaults to "V".
nObs	Number of observations. Used for testing significance of the canonical variate pairs.
alpha	Significance level at which the test will be evaluated.

Value

A list object with CCA information.

• corMat - The correlation matrix used for the analysis (unchanged).

• RCC - A list containing the raw correlation coefficients for both sets.

• SCC - A vector containing the sample canonical correlations between the variate pairs.

• R2 - A vector containing the squared sample canonical correlations between the variate pairs.

• prExVar - A dataframe containing the proportions of explained variance of the variate pairs.

• testRes - The result of the large sample test.

• additionalCorrelations - A list containing additional correlation matrices.

parhammarstrom/multivariateAnalysisVT2021 documentation built on March 22, 2021, 4:52 p.m.