read.snps: Read genotype matrix from file.
In stefanedwards/Siccuracy: Pipeline Package for AlphaImpute

Description Usage Arguments Details Value See Also Examples

read.snps is wrapper around scan that also converts into native matrix format.

1 2	read.snps(file, ncols = NULL, na = NA, what = integer(), extractIDs = TRUE, quiet = TRUE, ...)

`file`	Filename or connection.
`ncols`	Integer, number of columns in `file`, including ID column. When `NULL` (default), automagically detected with `get_ncols`.
`na`	Missing values; entries with this value are replaced with `NA`..
`what`	The typeof of data to be read, e.g. `integer()`, `numeric()`, or `character()`.
`extractIDs`	Logical, default `TRUE`, trim of first column and use as rownames.
`quiet`	logical: if `FALSE`, `scan()` will print a line, saying how many items have been read.
`...`	Passed on to `scan`.

Assumes a file format with no header and first column are IDs. If no ID column, use extractIDs = FALSE. Usually white-space delimted, but separator can be set with sep argument as per scan. For file format example see Siccuracy.

If ID columns contains alphabetical elements, use what=character(). This will however return a character matrix. Use e.g. storage.mode(m) <- 'integer' to convert to integer-based, keeping all other attributes.

Native matrix.

write.snps, typeof, get_ncols, scan, and Oxford, SHAPEIT, and VCF formats.

# Make test data
tmpfile <- tempfile()
write.snps(Siccuracy:::make.true(5, 10), tmpfile)
M <- read.snps(tmpfile)

# Process the genotypes in chunks:
# Setting mode to 'r' is very important to avoid resetting the pointer at file head!
ncols <- get_ncols(tmpfile)
f <- file(tmpfile, 'r') 
while (TRUE) {
  M <- read.snps(f, nlines=500, ncols=ncols)
  if (nrow(M) == 0) break
}
close(f)