exdex (version 1.2.3)

split_by_NAs: Divides data into parts that contain no missing values

Description

Splits the values in a numeric matrix column-wise into sequences of non-missing values.

Usage

split_by_NAs(x)

Value

A matrix containing a column for each run of non-missing values in

x. The number of rows is equal to the longest run of non-missing values in x and will therefore be at most nrow{x}. The matrix is padded with NA values at the end of each column, where necessary.

The returned object has an attribute called split_by_NAs_done

whose value is TRUE, so that in programming one can avoid calling

split_by_NAs more than once.

Arguments

x

A vector or matrix.

Details

For each column in x, split_by_NAs finds runs of values that contain no missing values and assigns them to a column in the matrix that is returned. Different columns are treated separately. If there are no missing values in a column then that column appears unmodified in the output matrix. Please see the Examples for illustrations.

Examples

# Create a simple numeric matrix and insert some NAs
x <- matrix(1:50, 10, 5)
x[c(3, 8), 1] <- NA
x[c(1:2, 5, 10), 3] <- NA
x[1:3, 4] <- NA
x[7:10, 5] <- NA
x

res <- split_by_NAs(x)
res

# An example of a character matrix
x <- matrix(c(letters, letters[1:18]), 11, 4)
x[c(1:2, 5:11), 2] <- NA
x[c(2:4, 6:11), 3] <- NA
x[1:10, 4] <- NA

res <- split_by_NAs(x)
res