fderiv: First derivatives of fitted GAM functions

In gavinsimpson/tsgam: Utilities for Working with GAMs Fitted to Time Series

Description

The first derivative of the smooth functions of a GAM model calculated using finite differences.

Usage

fderiv(model, ...)

## S3 method for class 'gam'
fderiv(model, newdata, term, n = 200, eps = 1e-07,
  unconditional = FALSE, ...)

## S3 method for class 'gamm'
fderiv(model, ...)

Arguments

model	A fitted GAM. Currently only models fitted by mgcv::gam() and mgcv::gamm() are supported.
...	Arguments that are passed to other methods.
newdata	a data frame containing the values of the model covariates at which to evaluate the first derivatives of the smooths.
term	character; vector of one or more terms for which derivatives are required. If missing, derivatives for all smooth terms will be returned.
n	integer; if newdata is missing the original data can be reconstructed from model and then n controls the number of values over the range of each covariate with which to populate newdata.
eps	numeric; the value of the finite difference used to approximate the first derivative.
unconditional	logical; if TRUE, the smoothing parameter uncertainty corrected covariance matrix is used, if available, otherwise the uncorrected Bayesian posterior covariance matrix is used.

Value

An object of class "fderiv" is returned.

Author(s)

Gavin L. Simpson

Examples

library("mgcv")
set.seed(2)
dat <- gamSim(1, n = 400, dist = "normal", scale = 2)
mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = "REML")

## first derivatives of all smooths...
fd <- fderiv(mod)

## ...and a selected smooth
fd2 <- fderiv(mod, term = "x1")

## Models with factors
set.seed(2)
dat <- gamSim(4, n = 400, dist = "normal", scale = 2)
mod <- gam(y ~ s(x0) + s(x1) + fac, data = dat, method = "REML")

## first derivatives of all smooths...
fd <- fderiv(mod)

## ...and a selected smooth
fd2 <- fderiv(mod, term = "x1")

gavinsimpson/tsgam documentation built on May 16, 2019, 10:11 p.m.