Effective sample size is applied even if not multinomial

[cgrandin]

Not sure if this was by design, this happens in the data file loading part of the tpl. Seems that the code should check if multinomial here:

		if(inp_nscaler(k) > 0)
			{
				for(i = 1; i <= n_A_nobs(k); i++ )
				{
					 tmp(i) = tmp(i)/sum(tmp(i)) * inp_nscaler(k);
				}
			}

[smartell]

Good point. I suppose the conditional statement should have 2 conditions

[cgrandin]

That's what I thought. I just sent you an email as well thinking you weren't looking on github.

[smartell]

The intention was to be able to scale the age-comp proportions to some other sample size when the data were entered in as proportions.