There are two plotting function here:
i) fitted_devianceIncr() plots the fitted model deviance components. This is useful for identifying observations with unusual y-values (given the current fitted distribution).
iii) model_devianceIncr_diff plots the difference of deviance increments from two fitted GAMLSS model. This function is useful if the GAIC and the residuals contradict each other. For example the GAIC is better for model 1 but the residuals look lot better for model 2. This can happens if the two distributions are better suited to fit different parts of the response distribution i.e. one model fits the center better but the other fits the tail better.
fitted_devianceIncr(obj, plot = TRUE, title, quan.val = 0.99,
annotate = TRUE, line.col = "steelblue4",
point.col = "darkblue", annot.col = "white",
newdata = NULL)
model_devianceIncr_diff(model1, model2, against = "index",
tol = 20, data, newdata)
return a plot
a GAMLSS fitted object
whether to create just the plot or save also the values with high deviance increment
a tittle if needed it.
The quantile values of the deviance increment from which the obsrevrvarion should be identify
whether to plot the quantile values above in the plot.
the colour of the line
the colour of the points
the colour of the annotation for the deviance increment plot
The first fitted GAMLSS model
The second fitted GAMLSS model
you can plot the deviance increment an index, the response or an x-variable
if the absolute value of deviance increment exceeds the tol the number of the observation is plotted
The data if can not be found from model1
evaluates the function in new data
Mikis Stasinopulos, Rober Rigby and Fernanda de Bastiani
The functions are diagnostic tools to check unusual observations in the response.
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.
Stasinopoulos, M.D., Kneib, T., Klein, N., Mayr, A. and Heller, G.Z., (2024). Generalized Additive Models for Location, Scale and Shape: A Distributional Regression Approach, with Applications (Vol. 56). Cambridge University Press.
(see also https://www.gamlss.com/).
gamlss
m1 <- gamlss(R~pb(Fl)+pb(A)+H+loc, data=rent, family=GA )
m2 <- gamlss(R~pb(Fl)+pb(A)+H+loc, data=rent, family=NO )
fitted_devianceIncr(m1)
model_devianceIncr_diff(m1,m2, against="Fl")
Run the code above in your browser using DataLab