# smoothing: Basis smoothing In asw221/dlm: Distributed Lag Models for Built Environment Applications

## Description

Constructs a set of basis vectors for distances between distributed lag points, and apply as a linear transformation of a concentration matrix.

## Usage

 ```1 2 3``` ```cr(x, Z, ...) sm(x, Z, ..., .fun = NULL) ```

## Arguments

 `x` a vector of values to construct the basis from. Missing values are not allowed `Z` a covariate matrix (or object that can be coerced to a `matrix`) to apply the linear basis transformation to. `length(x)` should be the same as `ncol(Z)` `...` arguments to be passed to `basis`

## Details

These functions are little more than convenient wrappers to the function `basis` and the `SmoothLag` class constructor. They are intended to simplify the task of specifying lag terms in a model `formula`. The functions compute a set of basis vectors for parameter `x` and applies this basis as a linear transformation of the covariate/concentration matrix parameter, `Z`. For example, if `cr` is used and `Z` is the identity matrix, the model fit will simply be the natural cubic spline of `x`.

Note that other basis extensions should always return an object that inherits from `SmoothLag`

## Value

An S4 object of class `SmoothLag`.

## Functions

• `cr`: natural cubic radial basis spline

• `sm`: user-defined smoothing

## References

Rupert D, Wand MP, & Carroll RJ (2003) Semiparametric Regression. New York: Cambridge University Press.

## See Also

`basis`, `SmoothLag`

## Examples

 ```1 2 3 4 5 6 7``` ```## load simulated data set and extract concentration matrix data (simdata) Conc <- simdata[, -(1:3)] # First columns are Y, Age, and Gender ## radial lag (distance) each concentration was measured at x <- seq(0.1, 10, length.out = ncol(Conc)) crb <- cr(x, Conc) ```

asw221/dlm documentation built on May 8, 2019, 5:59 p.m.