statList: Calculate statistics for levels of factors
In heplots: Visualizing Hypothesis Tests in Multivariate Linear Models

Description Usage Arguments Details Value Author(s) See Also Examples

statList provides a general method for calculating univariate or multivariate statistics for a matrix or data.frame stratified by one or more factors.

statList(X, factors, FUN, drop = FALSE, ...)

colMeansList(X, factors, drop = FALSE, ...)

covList(X, factors, drop = FALSE, ...)

`X`	A matrix or data frame containing the variables to be summarized
`factors`	A vector, matrix or data frame containing the factors for which `X` is to be summarized. If `factors` is not specified, the result is calculated for all of the data in `X`.
`FUN`	A function to be applied to the pieces of `X`, as split by `factors`.
`drop`	Logical, indicating whether empty levels of `factors` are to be dropped from the result.
`...`	Other arguments, passed to `FUN`.

statList is the general function. X is first split by factors, and FUN is applied to the result.

colMeansList and covList are just calls to statList with the appropriate FUN.

Returns a list of items corresponding to the unique elements in factors, or the interaction of factors. Each item is the result of applying FUN to that collection of rows of X. The items are named according to the levels in factors.

Michael Friendly

colMeans, termMeans

# grand means
statList(iris[,1:4], FUN=colMeans)
# species means
statList(iris[,1:4], iris$Species, FUN=colMeans)
# same
colMeansList(iris[,1:4], iris$Species)

# var-cov matrices, by species
covList(iris[,1:4], iris$Species)

# multiple factors
iris$Dummy <- sample(c("Hi","Lo"),150, replace=TRUE)
colMeansList(iris[,1:4], iris[,5:6])