Closed hfrick closed 1 year ago
closes #143
I've added to the documentation of the formula argument for flexsurvreg(). Let me know if you think that this gets the point across.
formula
flexsurvreg()
I've set the default for the new data argument to NULL so it doesn't change anything for flexsurvmix(). I figured that might need some more thinking.
data
NULL
flexsurvmix()
Are there other aspects the tests should cover?
I had the R CMD check fail when I checkout the repo, due to the vignette, but it passes with devtools::check(vignettes = FALSE).
devtools::check(vignettes = FALSE)
library(flexsurv) #> Loading required package: survival flexsurvreg(Surv(time, status) ~ ., data = lung, dist = "lognormal") #> Call: #> flexsurvreg(formula = Surv(time, status) ~ ., data = lung, dist = "lognormal") #> #> Estimates: #> data mean est L95% U95% se exp(est) #> meanlog NA 6.47e+00 3.52e+00 9.41e+00 1.50e+00 NA #> sdlog NA 9.97e-01 8.76e-01 1.13e+00 6.57e-02 NA #> inst 1.07e+01 2.00e-02 -1.37e-03 4.14e-02 1.09e-02 1.02e+00 #> age 6.26e+01 -1.44e-02 -3.43e-02 5.41e-03 1.01e-02 9.86e-01 #> sex 1.38e+00 4.76e-01 1.20e-01 8.32e-01 1.82e-01 1.61e+00 #> ph.ecog 9.58e-01 -5.31e-01 -9.38e-01 -1.25e-01 2.08e-01 5.88e-01 #> ph.karno 8.20e+01 -1.44e-02 -3.74e-02 8.60e-03 1.17e-02 9.86e-01 #> pat.karno 7.96e+01 8.12e-03 -5.00e-03 2.12e-02 6.70e-03 1.01e+00 #> meal.cal 9.29e+02 2.37e-04 -1.66e-04 6.39e-04 2.06e-04 1.00e+00 #> wt.loss 9.72e+00 9.84e-03 -2.61e-03 2.23e-02 6.35e-03 1.01e+00 #> L95% U95% #> meanlog NA NA #> sdlog NA NA #> inst 9.99e-01 1.04e+00 #> age 9.66e-01 1.01e+00 #> sex 1.13e+00 2.30e+00 #> ph.ecog 3.91e-01 8.83e-01 #> ph.karno 9.63e-01 1.01e+00 #> pat.karno 9.95e-01 1.02e+00 #> meal.cal 1.00e+00 1.00e+00 #> wt.loss 9.97e-01 1.02e+00 #> #> N = 167, Events: 120, Censored: 47 #> Total time at risk: 51759 #> Log-likelihood = -837.4184, df = 10 #> AIC = 1694.837
Created on 2022-10-13 by the reprex package (v2.0.1)
That looks good to me, thank you! I copped out of trying to implement it in flexsurvmix, but added an error message and documentation about it not being supported there.
closes #143
I've added to the documentation of the
formula
argument forflexsurvreg()
. Let me know if you think that this gets the point across.I've set the default for the new
data
argument toNULL
so it doesn't change anything forflexsurvmix()
. I figured that might need some more thinking.Are there other aspects the tests should cover?
I had the R CMD check fail when I checkout the repo, due to the vignette, but it passes with
devtools::check(vignettes = FALSE)
.Created on 2022-10-13 by the reprex package (v2.0.1)