Fit Frechet distribution to non-censored data

Source: R/salsa-fit.R

Fit Frechet distribution to non-censored data

fitdist_fr(x, method = "mme", min_shape = 1, order = c(1, 2),
  central = FALSE, absolute = FALSE, na.rm = FALSE,
  verbose = FALSE, ...)

Arguments

x

numeric vector
method

fitting method, argument passed to fitdistrplus::fitdist(), by default "mme" moment matching estimation.
min_shape

numeric value restricting the minimum shape otherwise output is set to NULL.
order, central, absolute, na.rm

arguments passed to moments::moment(). These values do not typically need to be customized.
verbose

logical indicating whether to print verbose output.
...

additional arguments are passed to fitdistrplus::fitdist().
param_fr_wei

data.frame of parameter bounds, output from params_fr_wei(), which defines the start, upper, and lower bounds for each of the five parameters.

Value

object of class "fitdist" described in fitdistrplus::fitdist(). If any error occurs during the fit, a NULL is returned.

Details

This function is a simple wrapper around fitdistrplus::fitdist() for the Frechet distribution, using "actuar::dinvweibull". Most customizations are performed by the corresponding support function, for example when method="mme" the distribution fit function is fitdistrplus::mmedist(), which then calls stats::optim(). Arguments are passed to these functions using ... where applicable.

See also

Other SALSA distribution fit functions: dfr_wei, fitdist_fr_wei, params_fr_wei

Examples

data(oz2_numi_per_cell);
x <- oz2_numi_per_cell$count[oz2_numi_per_cell$count >= 16];
fit_fr <- fitdist_fr(x);
# print coefficients
print(coef(fit_fr));
#>     shape     scale 
#>  2.003168 25.184591