Allows the user to specify a loess fit within a GAMLSS model. This function is similar to the lo function in the gam implementation of package gam see Chambers and Hastie (1991).
The function vis.lo() allows plotting the results.
lo(formula, control = lo.control(...), ...)
lo.control(span = 0.75, enp.target = NULL,
degree = 2, parametric = FALSE, drop.square = FALSE,
normalize = TRUE, family = c("gaussian", "symmetric"),
method = c("loess", "model.frame"),
surface = c("interpolate", "direct"),
statistics = c("approximate", "exact", "none"),
trace.hat = c("exact", "approximate"),
cell = 0.2, iterations = 4,iterTrace = FALSE, ...)
vis.lo(obj, se=-1, rug = FALSE, partial.resid = FALSE,
col.term = "darkred", col.shaded = "gray",
col.res = "lightblue", col.rug = "gray", lwd.term = 1.5,
cex.res = 1, pch.res = par("pch"),
type = c("persp", "contour"), col.surface = "gray",
nlevels = 30, n.grid = 30, image = TRUE, ...)a loess object is returned.
a formula specifying the explanatory variables
a control to be passed to the loess function
extra arguments
the number of observations in a neighbourhood. This is the smoothing parameter for a loess fit.
an alternative way to specify span, as the approximate equivalent number degrees of freedom to be used. See also the help file of the R function loess. For consistency with the older version of lo the effective degrees of freedom df can be also specified instead of span, e.g. df=5
the degree of local polynomial; can be 1 or 2. See also the help file of loess
should any terms be fitted globally rather than locally? See the help file of loess
for fits with more than one predictor and degree=2, should the quadratic term be dropped for particular predictors?. See also help file of loess
should the predictors be normalized to a common scale if there is more than one? See the help file of loess
if "gaussian" fitting is by least-squares, and if "symmetric" a re-descending M estimator is used with Tukey's biweight function. See the help file of loess
fit the model or just extract the model frame. See the help file of loess
should the fitted surface be computed exactly or via interpolation from a kd tree? See also
the help file of loess.control
should the statistics be computed exactly or approximately? See the help file of loess.control
should the trace of the smoother matrix be computed exactly or approximately? See the help file of loess.control
if interpolation is used this controls the accuracy of the approximation via the maximum number of points in a cell in the kd tree. See the help file of loess.control
the number of iterations used in robust fitting. See the help file of loess.control
logical (or integer) determining if tracing information during the robust iterations (iterations>= 2) is produced. See the help file of loess.control
an lowss object fitted within gamlss
if se>0 then standard errors surfaces are drawn in the 3-dimensional plot. Set se at the required level i.e se=1.96 will be an approximated 95% CI.
whether to plot a rug in the plot
whether to plot the partial residuals
the colour of the line of fitted term
the shading of standard
the shading of standard error intervals
the colour of partial residuals
the colour of the rug
the width of the line
The character for the partial residuals
The type of the plot if the x's are two dimensional
the colour of the fitted surface
the number of levels used in cotour() plot.
The number of points to evaluate the surface
whether to use image() or just contour
Mikis Stasinopoulos d.stasinopoulos@londonmet.ac.uk, Bob Rigby, (The original lo() function was based on the Trevor Hastie's S-plus lo() function. See also the documentation of the loess function for the authorship of the function.
In this version the first argument is a formula NOT a list as in the previous one
Note that lo itself does no smoothing; it simply sets things up for the function gamlss.lo() which is used by the backfitting function gamlss.add().
Chambers, J. M. and Hastie, T. J. (1991). Statistical Models in S, Chapman and Hall, London.
Rigby, R. A. and Stasinopoulos D. M. (2005). Generalized additive models for location, scale and shape,(with discussion), Appl. Statist., 54, part 3, pp 507-554.
Rigby, R. A., Stasinopoulos, D. M., Heller, G. Z., and De Bastiani, F. (2019) Distributions for modeling location, scale, and shape: Using GAMLSS in R, Chapman and Hall/CRC. An older version can be found in https://www.gamlss.com/.
Stasinopoulos D. M. Rigby R.A. (2007) Generalized additive models for location scale and shape (GAMLSS) in R. Journal of Statistical Software, Vol. 23, Issue 7, Dec 2007, https://www.jstatsoft.org/v23/i07/.
Stasinopoulos D. M., Rigby R.A., Heller G., Voudouris V., and De Bastiani F., (2017) Flexible Regression and Smoothing: Using GAMLSS in R, Chapman and Hall/CRC.
(see also https://www.gamlss.com/).
cs, random,
# fitting a loess curve with span=0.4 plus the a quarterly effect
aids1<-gamlss(y~lo(~x,span=0.4)+qrt,data=aids,family=PO) #
term.plot(aids1, page=1)
if (FALSE) {
r1 <- gamlss(R~lo(~Fl)+lo(~A), data=rent, family=GA)
term.plot(r1, pages=1)
vis.lo(getSmo(r1, which=1), partial=T)
r2 <- gamlss(R~lo(~Fl+A), data=rent, family=GA)
term.plot(r2, pages=1)
vis.lo(getSmo(r2, which=1))
vis.lo(getSmo(r2, which=1), se=1.97)
vis.lo(getSmo(r2, which=1), partial.res=T)
}
Run the code above in your browser using DataLab