reconstruct_four function

Reconstruct a set of reformatted four-fold tables

Sometimes, fourfold tables are reformatted by replacing rows or columns by marginal totals. This makes it impossible to use them straight away for statistical tests like Fisher's exact test. But with that knowledge, the missing values can easily be restored. The reconstruct_four

function uses a set of such reduced tables, stored row-wise in a matrix or a data frame, and rebuilds the two reformatted cells when they were replaced by marginal totals.

reconstruct_four(dat, idx_marginals = NULL, colnames_add = NULL)

Arguments

• dat: integer matrix or data frame with exactly two columns; each row represents the first column of a 2x2 matrix for which the other two values are to be computed and appended to dat as two new columns; real numbers will be coerced to integer.
• idx_marginals: integer vector of exactly two values or NULL (the default) indicating the columns of dat that contain the marginal totals; if NULL, the last two columns are used.
• colnames_add: character vector of exactly two unique character strings or NULL (the default), which contains the desired headers of the new (reconstructed) columns of the input; if NULL, the headers of the marginal totals are used.

Returns

An integer data frame with four columns.

Examples

X1 <- c(4, 2, 2, 14, 6, 9, 4, 0, 1)
X2 <- c(0, 0, 1, 3, 2, 1, 2, 2, 2)
N1 <- rep(148, 9)
N2 <- rep(132, 9)

df1 <- data.frame(X1, X2, N1, N2)
reconstruct_four(df1, colnames_add = c("Y1", "Y2"))
# same as reconstruct_four(df1, c(3, 4), c("Y1", "Y2"))

df2 <- data.frame(X1, N1, X2, N2)
reconstruct_four(df2, c(2, 4), c("Y1", "Y2"))