`registerDistributions` type-checking does not allow default values for `x`

[danielturek]

When defining the density function of a custom distribution, the type-checking that takes place in registerDistributions does not allow a default value for x. During the type-checking, we require an identical to pass for the return type defined by the density function (in the dist <- nimbleFunction(x = ...) of the argument list), and the returnType statement of the generation function. If you provide a default value of x in the density function, this identical check fails. Example below:

library(nimble)

ddist <- nimbleFunction(
    run = function(x = double(0, default = 0), log = integer(default = 1)) {
        returnType(double())
        return(0)
    }
)

rdist <- nimbleFunction(
    run = function(n = integer()) {
        returnType(double())
        return(0)
    }
)

registerDistributions('ddist')

## Error in FUN(X[[i]], ...) : 
##   checkDistributionFunctions: random generation function `rdist` is missing `returnType` or `returnType` 
##   does not match the type of the `x` argument to the corresponding density function.

[paciorek]

I think I would argue that instead we should not allow default arguments for user-defined distributions (by analogy with standard distributions).

@danielturek what do you think about me adding a check that errors out if there is a default?

[danielturek]

@paciorek Yes, that makes more sense, stepping back and looking at this again. I'm on board with your suggestion.

[paciorek]

Ok, I've added an error trap in checkDistributionFunctions in branch fix_1345.

[paciorek]

Fixed by PR #1371