rational_biweight_fac: Tukey's biweight of rational function models
In weiyaw/blackbox: A SMC-SA Algorithm for Constrained Optimisation

Description Usage Arguments Details Value

Produce a function that calculates the Tukey's biweight of a set of rational function models.

1	rational_biweight_fac(y, x, dim_f, dim_g, cst = 4.685)

`y`	a response vector. (num vec)
`x`	a predictor vector. (num vec)
`dim_f`	the number of coefficients in the numerator. (num)
`dim_g`	the number of coefficients in the denominator, excluding the constant term. (num)
`cst`	the constant in Tukey's biweight. (num)

The explicit formula used is rho(x) = (c^2)/6 * (1-(1-(x/c)^2)^3) if x < |c|, or rho(x) = 1 if x > |c|, where c is an arbitrary constant. The sum of rho(y_i - r(x_i)), where r(.) is the rational function, is then minimised. The produced function can be minimised using SMC-SA, multiSA or CEPSO.

a function that takes in a matrix of rational function coefficients (numerator precede denominator, in ascending order), each column corresponds to a rational function, and returns a vector Tukey's biweight.

weiyaw/blackbox documentation built on June 7, 2019, 5:12 a.m.