plot.maxlogL: Plot Residual Diagnostics for an 'maxlogL' Object

In Jaimemosg/EstimationTools: Maximum Likelihood Estimation for Probability Functions from Data Sets

Plot Residual Diagnostics for an maxlogL Object

Description

Provides plots of Cox-Snell, martingale Randomized quantile residuals.

Usage

## S3 method for class 'maxlogL'
plot(
  x,
  type = c("rqres", "cox-snell", "martingale", "right-censored-deviance"),
  parameter = NULL,
  which.plots = NULL,
  caption = NULL,
  xvar = NULL,
  ...
)

Arguments

x	a maxlogL object.
type	a character with the type of residuals to be plotted. The default value is type = "rqres", which is used to compute the normalized randomized quantile residuals.
parameter	which parameter fitted values are required for type = "rqres". The default is the first one defined in the pdf,e.g, in dnorm, the default parameter is mean.
which.plots	a subset of numbers to specify the plots. See details for further information.
caption	title of the plots. If caption = NULL, the default values are used.
xvar	an explanatory variable to plot the residuals against.
...	further parameters for the plot method.

Details

If type = "rqres", the available subset is 1:4, referring to:

• 1. Quantile residuals vs. fitted values (Tukey-Ascomb plot)

• 2. Quantile residuals vs. index

• 3. Density plot of quantile residuals

• 4. Normal Q-Q plot of the quantile residuals.

Value

Returns specified plots related to the residuals of the fitted maxlogL model.

Author(s)

Jaime Mosquera Gutiérrez, [e-mail]

Examples

library(EstimationTools)

#----------------------------------------------------------------------------
# Example 1: Quantile residuals for a normal model
n <- 1000
x <- runif(n = n, -5, 6)
y <- rnorm(n = n, mean = -2 + 3 * x, sd = exp(1 + 0.3* x))
norm_data <- data.frame(y = y, x = x)

# It does not matter the order of distribution parameters
formulas <- list(sd.fo = ~ x, mean.fo = ~ x)
support <- list(interval = c(-Inf, Inf), type = 'continuous')

norm_mod <- maxlogLreg(formulas, y_dist = y ~ dnorm, support = support,
                       data = norm_data,
                       link = list(over = "sd", fun = "log_link"))

# Quantile residuals diagnostic plot
plot(norm_mod, type = "rqres")
plot(norm_mod, type = "rqres", parameter = "sd")

# Exclude Q-Q plot
plot(norm_mod, type = "rqres", which.plots = 1:3)


#----------------------------------------------------------------------------
# Example 2: Cox-Snell residuals for an exponential model
data(ALL_colosimo_table_4_1)
formulas <- list(scale.fo = ~ lwbc)
support <- list(interval = c(0, Inf), type = 'continuous')

ALL_exp_model <- maxlogLreg(
  formulas,
  fixed = list(shape = 1),
  y_dist = Surv(times, status) ~ dweibull,
  data = ALL_colosimo_table_4_1,
  support = support,
  link = list(over = "scale", fun = "log_link")
)

summary(ALL_exp_model)
plot(ALL_exp_model, type = "cox-snell")
plot(ALL_exp_model, type = "right-censored-deviance")

plot(ALL_exp_model, type = "martingale", xvar = NULL)
plot(ALL_exp_model, type = "martingale", xvar = "lwbc")


#----------------------------------------------------------------------------

Jaimemosg/EstimationTools documentation built on Oct. 23, 2023, 10 a.m.