Should year columns be integers?

[bpbond]

Currently (see #802 and #820 ) the driver checks that all year columns are numeric. Should they forced to be integers? To do this right, if yes we should change many definitions in constants.R to e.g.

HISTORICAL_YEARS <- 1971L:2010L

[rplzzz]

That would appear not to be the case:

> is.integer(1971:2010)
[1] TRUE

It seems that : and seq prefer to create integers whenever their arguments are integer-valued numerics. For example:

> is.integer(seq(1971,1972))
[1] TRUE
> is.integer(c(1971,1972))
[1] FALSE

[bpbond]

Oh, huh, interesting–good for you for checking.