stats (version 3.3)

numericDeriv: Evaluate Derivatives Numerically

Description

numericDeriv numerically evaluates the gradient of an expression.

Usage

numericDeriv(expr, theta, rho = parent.frame(), dir = 1.0)

Arguments

expr

The expression to be differentiated. The value of this expression should be a numeric vector.

theta

A character vector of names of numeric variables used in expr.

rho

An environment containing all the variables needed to evaluate expr.

dir

A numeric vector of directions to use for the finite differences.

Value

• The value of eval(expr, envir = rho) plus a matrix attribute called gradient. The columns of this matrix are the derivatives of the value with respect to the variables listed in theta.

Details

This is a front end to the C function numeric_deriv, which is described in Writing R Extensions.

The numeric variables must be of type real and not integer.

Examples

<testonly>od <- options(digits = 4)</testonly>
myenv <- new.env()
assign("mean", 0., envir = myenv)
assign("sd", 1., envir = myenv)
assign("x", seq(-3., 3., len = 31), envir = myenv)
numericDeriv(quote(pnorm(x, mean, sd)), c("mean", "sd"), myenv)
<testonly>options(od)</testonly>