Generate a sum of squares cost function for optimization

This is an internal helper function that generates a sum of squares cost function for any given r* function (e.g. rnorm, rlognorm). The resulting function is intended to be used by an optim call for fitting quantiles to distribution parameters.

generate_cost_function(func)

Arguments

func	A distribution function.

Value

A function.

See also

Other distribution fitting functions: combine_lognorm_trunc(), combine_lognorm(), combine_norm(), fit_capabilities_geomean(), fit_capabilities(), fit_lognorm_trunc(), fit_lognorm(), fit_norm_trunc(), fit_pois(), fit_scenarios_geomean(), fit_scenarios(), fit_threat_communities(), lognormal_to_normal(), normal_to_lognormal()

Examples

generate_cost_function(stats::qlnorm)
#> function (x, quant, est, ...) 
#> {
#>     x1 <- x[1]
#>     x2 <- x[2]
#>     if (x1 < 0 | x2 < 0) 
#>         return(NA)
#>     sum(((rlang::get_expr(func))(quant, x1, x2, ...) - est)^2)
#> }
#> <bytecode: 0x7feca5b7d838>
#> <environment: 0x7fec871cca90>