powered by
Transforms a matrix via the specified function.
transform(x, type, par, deriv_order)
Returns an object of class derivs.
derivs
numeric matrix to be transformed.
string, specifies the transformation function. Available are:
`identity`: \(f(x)=x\).
`exp`: \(f(x)=\exp\{x\}\).
`log`: \(f(x)=\log\{x\}\).
`glogit`: \(f(x)=\log\{(-x + min)/(x - max)\), where par=c(min, max).
par=c(min, max)
`glogitinv`: \(f(x)=\exp\{x\} \cdot (max + min)/(1 + \exp\{x\}) \), where par=c(min, max).
`inv`: \(f(x)=\frac{1}{x}\).
`pnorm`: \(f(x)=\Phi(x)\).
`qnorm`: \(f(x)=\Phi^{-1}(x)\).
`mexp`: \(f(x)=-\exp\{x\}\).
`zeta`: \(f(x)=\log\{2 \cdot \Phi(x)\}\).
`constant`: \(f(x)=c\).
`chainrule_utility`: \(f(x)=f'(x)=f''(x)=f'''(x)=f''''(x)\).
onemx: \(1-x\)
numeric vector, additional parameters, e.g. min and max for glogit.
glogit
integer; maximum order of derivative. Available are 0,2 and 4.
0
2
4
Takes the numeric matrix x as an input for the function specified by type and evaluates it together with the derivatives.
type
A<-matrix(c(1:9)/10, ncol=3) A_mat<-list2derivs(list(A, A^0, A^2, A^3, A^4), deriv_order=4) transform(x=transform(x = A, type="exp", par=0, deriv_order=4), type="log", deriv_order=4, par = 0)
Run the code above in your browser using DataLab