Update the dataset path in NONMEM model code

Replaces the dataset path in the first $DATA record with a new path, while preserving every other element: any options on the $DATA line (IGNORE=@, ACCEPT=(...), REWIND, RECORDS=N, ...), any inline comment on that line, and all other records in the model code.

Usage

update_nonmem_data(code, path)

Arguments

code

NONMEM model code. Either a single character string (possibly with embedded newlines) or a character vector of lines.

path

path to the new dataset.

Value

The updated model code, in the same shape as code: a single string if code was a string, a character vector otherwise.

Details

The match is case-insensitive and also accepts the abbreviated $DAT record name. Dataset paths in the original record may be unquoted or wrapped in single or double quotes. If the new path contains whitespace, it is wrapped in single quotes so NONMEM parses it as one token.

Examples

code <- "$PROB Test\n$DATA old.csv IGNORE=@ ACCEPT=(DV.GT.0)\n$INPUT ID TIME DV"
cat(update_nonmem_data(code, "new.csv"))
#> $PROB Test
#> $DATA new.csv IGNORE=@ ACCEPT=(DV.GT.0)
#> $INPUT ID TIME DV