For most crosses, the result is either NULL
(indicating no additional covariates are needed) or a
matrix with a single column containing sex indicators (1 for males
and 0 for females).
For an intercross, we also consider cross direction. There are
four cases: (1) All male or all female but just one direction: no
covariate; (2) All female but both directions: covariate indicating
cross direction; (3) Both sexes, one direction: covariate
indicating sex; (4) Both sexes, both directions: a covariate
indicating sex and a covariate that is 1 for females from the
reverse direction and 0 otherwise.