Skip to contents

creates a list of class ktab from a list of data frames

Usage

ktab.list.df(obj, rownames = NULL, colnames = NULL, tabnames = NULL, 
    w.row = rep(1, nrow(obj[[1]])), w.col = lapply(obj, function(x) 
    rep(1 / ncol(x), ncol(x))))

Arguments

obj

a list of data frame

rownames

the names of the K-tables rows (otherwise, the row names of the arrays)

colnames

the names of the K-tables columns (otherwise, the column names of the arrays)

tabnames

the names of the arrays of the K-tables (otherwise, the names of the obj if they exist, or else "Ana1", "Ana2", ...)

w.row

a vector of the row weightings in common with all the arrays

w.col

a list of the vector of the column weightings for each array

Details

Each element of the initial list have to possess the same names and row numbers

Author

Daniel Chessel
Anne-Béatrice Dufour anne-beatrice.dufour@univ-lyon1.fr

Value

returns a list of class ktab. See ktab

Examples

data(jv73)
l0 <- split(jv73$morpho, jv73$fac.riv)
l0 <- lapply(l0, function(x) data.frame(t(scalewt(x))))
#> Warning: Variables with null variance not standardized.
#> Warning: Variables with null variance not standardized.
#> Warning: Variables with null variance not standardized.
kta <- ktab.list.df(l0)
kplot(sepan(kta[c(2, 5, 7, 10)]), perm = TRUE)
#> Error in s.label(dfxy = sepan(kta[c(2, 5, 7, 10)])$Co, labels = sepan(kta[c(2,     5, 7, 10)])$TC[, 2], facets = sepan(kta[c(2, 5, 7, 10)])$TC[,     1], xax = 1, yax = 2, plot = FALSE, storeData = TRUE, pos = -3,     psub = list(position = "bottomright"), samelimits = FALSE): non convenient selection for dfxy (can not be converted to dataframe)