Skip to contents

Extract model coefficients from a fitted mixturecure object

Source: R/coef.R
coef.mixturecure.Rd

coef.mixturecure is a generic function which extracts the model coefficients from a fitted mixturecure model object fit using curegmifs, cureem, cv_curegmifs, or cv_cureem.

Usage

# S3 method for class 'mixturecure'
coef(object, model_select = "AIC", ...)

Arguments

object

a mixturecure object resulting from curegmifs, cureem, cv_curegmifs, or cv_cureem.

model_select

either a case-sensitive parameter for models fit using curegmifs or cureem or any numeric step along the solution path can be selected. The default is model_select = "AIC" which calculates the predicted values using the coefficients from the model achieving the minimum AIC. The complete list of options are:

  • "AIC" for the minimum AIC (default).

  • "mAIC" for the minimum modified AIC.

  • "cAIC" for the minimum corrected AIC.

  • "BIC", for the minimum BIC.

  • "mBIC" for the minimum modified BIC.

  • "EBIC" for the minimum extended BIC.

  • "logLik" for the step that maximizes the log-likelihood.

  • n where n is any numeric value from the solution path.

This option has no effect for objects fit using cv_curegmifs or cv_cureem.

...

other arguments.

Value

rate

estimated rate parameter when fitting a Weibull or exponential mixture cure model.

shape

estimated shape parameter when fitting a Weibull mixture cure model.

b0

estimated intercept for the incidence portion of the mixture cure model.

beta_inc

the vector of coefficient estimates for the incidence portion of the mixture cure model.

beta_lat

the vector of coefficient estimates for the latency portion of the mixture cure model.

p_uncured

a vector of probabilities from the incidence portion of the fitted model representing the P(uncured).

See also

curegmifs, cureem, summary.mixturecure, plot.mixturecure, predict.mixturecure

Examples

library(survival)
withr::local_seed(1234)
temp <- generate_cure_data(n = 100, j = 10, n_true = 10, a = 1.8)
training <- temp$training
fit <- curegmifs(Surv(Time, Censor) ~ .,
  data = training, x_latency = training,
  model = "weibull", thresh = 1e-4, maxit = 2000,
  epsilon = 0.01, verbose = FALSE
)
coef(fit)
#> $rate
#> [1] 3.584565
#> 
#> $shape
#> [1] 1.197495
#> 
#> $b0
#> [1] 0.3670091
#> 
#> $beta_inc
#>    U1    U2    X1    X2    X3    X4    X5    X6    X7    X8    X9   X10 
#>  0.00  0.07  1.46 -1.23  1.22 -1.33  0.04 -1.37 -0.90  0.36 -0.07  1.77 
#> 
#> $beta_lat
#>    U1    U2    X1    X2    X3    X4    X5    X6    X7    X8    X9   X10 
#>  0.37  0.34  1.24  1.07  0.39  1.11 -0.84  0.04  0.71  0.38 -0.45 -1.20 
#>