Closed fa1999abdi closed 3 months ago
bblodfon
commented
4 months ago @fa1999abdi Can you please copy your code, run reprex::reprex(), and then copy-paste the result in the above post? Add the library versions as well with devtools::session_info(). Then add specific questions/comments on the code, ie what does work, what doesn't? Where you don't understand what is happening exactly? what you need to do that is not there? All that would help me help you.
fa1999abdi
commented
4 months ago Hi John I would like to share my code with you while maintaining its privacy. Can I send you the Reprex file in a secure environment?
On Sun, May 12, 2024 at 1:52 PM John Zobolas @.***> wrote:
Can you please copy your code, run reprex::reprex(), and then copy-paste the result in the above post? Add the library versions as well with devtools::session_info(). Then add specific questions/comments on the code, ie what does work, what doesn't? Where you don't understand what is happening exactly? what you need to do that is not there? All that would help me help you.
— Reply to this email directly, view it on GitHub https://github.com/mlr-org/mlr3proba/issues/381#issuecomment-2106197646, or unsubscribe https://github.com/notifications/unsubscribe-auth/BF3TSUX6NIRW2TZVEOIQHMDZB47G7AVCNFSM6AAAAABHSV6MDKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCMBWGE4TONRUGY . You are receiving this because you authored the thread.Message ID: @.***>
bblodfon
commented
4 months ago Hi @fa1999abdi, this is an open source research project and so I won't spend any time looking at private code. These issues are also for meant for the community so that others that might have similar problems benefit from this.
fa1999abdi
commented
4 months ago I want to make a prediction of the survival time on the test data using the gradient boosting learner and draw the survival graph for the test data. The questions next to the codes have been commented on.
library(mlr3verse)
#> Loading required package: mlr3
library(mlr3proba)
library(xgboost)
library(tidyverse)
library(survival)
library(mlr3)
library(mlr3proba)
library(mlr3learners)
library(mlr3filters)
library(reprex)
set.seed(42)
train_indxs = sample(seq_len(nrow(veteran)), 100)
task = as_task_surv(x = veteran, time = 'time', event = 'status')
poe = po('encode')
task = poe$train(list(task))[[1]]
set.seed(42)
part = partition(task, ratio = 0.8)
#I have tuned hyperparameters the learner with auto-tuning on train data.
learner_xgboost = lrn("surv.xgboost", eta = to_tune(1e-4, 1),gamma = to_tune(1e-4, 1),
max_depth = to_tune(1,35), min_child_weight = to_tune(0, 10))
learner_xgboost$param_set$search_space()
#> <ParamSet>
#> id class lower upper nlevels default value
#> 1: eta ParamDbl 1e-04 1 Inf <NoDefault[3]>
#> 2: gamma ParamDbl 1e-04 1 Inf <NoDefault[3]>
#> 3: max_depth ParamInt 1e+00 35 35 <NoDefault[3]>
#> 4: min_child_weight ParamDbl 0e+00 10 Inf <NoDefault[3]>
set.seed(456)
at_xgboost = auto_tuner(
tuner = tnr("random_search", batch_size = 50),
learner = learner_xgboost,
resampling = rsmp("cv", folds = 3),
measure = msr("surv.cindex"),
terminator = trm("evals", n_evals = 50))
modelxgboost <- at_xgboost$train(task, part$train)
#> INFO [16:07:00.462] [bbotk] Starting to optimize 4 parameter(s) with '<OptimizerRandomSearch>' and '<TerminatorEvals> [n_evals=50, k=0]'
#> INFO [16:07:00.513] [bbotk] Evaluating 50 configuration(s)
#> INFO [16:07:00.537] [mlr3] Running benchmark with 150 resampling iterations
#> INFO [16:07:00.603] [mlr3] Applying learner 'surv.xgboost' on task 'veteran' (iter 1/3)
#> INFO [16:07:00.671] [mlr3] Applying learner 'surv.xgboost' on task 'veteran' (iter 2/3)
#> ...
#> INFO [16:07:08.213] [mlr3] Applying learner 'surv.xgboost' on task 'veteran' (iter 2/3)
#> INFO [16:07:08.261] [mlr3] Applying learner 'surv.xgboost' on task 'veteran' (iter 3/3)
#> INFO [16:07:08.358] [mlr3] Finished benchmark
#> INFO [16:07:09.130] [bbotk] Result of batch 1:
#> INFO [16:07:09.134] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [16:07:09.134] [bbotk] 0.69240361 0.804262526 3 8.8348667 0.6641731 0 0
#> INFO [16:07:09.134] [bbotk] 0.06123976 0.987442239 26 1.5563251 0.6863813 0 0
#> INFO [16:07:09.134] [bbotk] 0.86951133 0.623217709 32 6.2168556 0.6887658 0 0
#> INFO [16:07:09.134] [bbotk] 0.87223711 0.830520471 10 3.9803411 0.7069063 0 0
#> INFO [16:07:09.134] [bbotk] 0.43205868 0.889244607 7 0.6064467 0.6775195 0 0
#> ...
#> INFO [16:07:09.134] [bbotk] 0.34691562 0.538365898 6 0.1004618 0.6523724 0 0
#> INFO [16:07:09.134] [bbotk] 0.26071728 0.936020394 28 4.2226101 0.6894068 0 0
#> INFO [16:07:09.134] [bbotk] 0.63756977 0.646424119 25 8.1569937 0.6821928 0 0
#> INFO [16:07:09.134] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [16:07:09.134] [bbotk] runtime_learners uhash
#> INFO [16:07:09.134] [bbotk] 0.12 ee4981c6-bd07-4919-8147-4ca021e621c0
#> INFO [16:07:09.134] [bbotk] 0.09 aaa67e14-d32b-492a-96bc-5246f7a8d993
#> INFO [16:07:09.134] [bbotk] 0.09 9aa90106-0b83-4894-95b4-ed67b40c52dc
#> INFO [16:07:09.134] [bbotk] 0.09 0e4fe10e-c165-498c-9d32-b597942cbf43
#> INFO [16:07:09.134] [bbotk] 0.10 10d1964b-b56e-4cac-b3b9-7baf63e913a3
#> INFO [16:07:09.134] [bbotk] 0.07 9df279e3-bae1-4fef-837c-c03491abc69b
#> INFO [16:07:09.134] [bbotk] 0.35 eb095987-6221-4b6a-a3d8-45eb78504aa9
#> INFO [16:07:09.134] [bbotk] 0.09 7b097928-ef11-4cdd-ac24-f53931ebda25
#> INFO [16:07:09.134] [bbotk] 0.09 58b02576-bfe1-4b6e-9945-4974b113bfc4
#> INFO [16:07:09.134] [bbotk] 0.09 3e75a77c-5f84-4c4f-bab4-c87446cea6a7
#> INFO [16:07:09.134] [bbotk] 0.16 907b12c8-d61e-45dd-9092-f904786106d1
#> INFO [16:07:09.134] [bbotk] 0.10 94231a85-7ad5-4713-bd24-0e642bd0847f
#> INFO [16:07:09.134] [bbotk] 0.09 eb417ce3-e597-43b1-a820-7909a63ca1e3
#> INFO [16:07:09.134] [bbotk] 0.09 9d4de986-135f-4a45-b311-ad1b1e251e89
#> ...
#> INFO [16:07:09.134] [bbotk] 0.10 d14b06be-ad6c-4201-a377-1ebe01af56a6
#> INFO [16:07:09.134] [bbotk] 0.09 33564a8b-a664-461c-a614-8490906519de
#> INFO [16:07:09.134] [bbotk] 0.08 6ecac5d0-155a-4bc9-ba1e-010f5dd2575c
#> INFO [16:07:09.134] [bbotk] runtime_learners uhash
#> INFO [16:07:09.144] [bbotk] Finished optimizing after 50 evaluation(s)
#> INFO [16:07:09.145] [bbotk] Result:
#> INFO [16:07:09.146] [bbotk] eta gamma max_depth min_child_weight learner_param_vals x_domain
#> INFO [16:07:09.146] [bbotk] 0.8722371 0.8305205 10 3.980341 <list[8]> <list[4]>
#> INFO [16:07:09.146] [bbotk] surv.cindex
#> INFO [16:07:09.146] [bbotk] 0.7069063
#Recursive Feature Elimination with Cross Validation
instance = fsi(
task =(task),#How to use train data(part$train)?
#How can I use tuned Lerner in the previous step to Select the variable?
learner = lrn("surv.xgboost"),
resampling = rsmp("cv", folds = 6),
measures = msr("surv.cindex"),
terminator = trm("none"),
)
optimizer$optimize(instance) #Why does it give an error?
#> Error in eval(expr, envir, enclos): object 'optimizer' not found
#I also want to subset the task to the optimal feature set
#and again train the learner.
task$select(instance$result_feature_set)
#> Error in .__Task__select(self = self, private = private, super = super, : Assertion on 'cols' failed: Must be of type 'character', not 'NULL'.
learner$train(task,part$train)
#> Error in eval(expr, envir, enclos): object 'learner' not found
#And finally, I use the trained model to predict the survival time on test data,
#which will encounter an error.
p =learner$predict(task, part$test)
#> Error in eval(expr, envir, enclos): object 'learner' not found
devtools :: session_info()
#> ─ Session info ───────────────────────────────────────────────────────────────
#> setting value
#> version R version 4.3.2 (2023-10-31 ucrt)
#> os Windows 10 x64 (build 19045)
#> system x86_64, mingw32
#> ui RTerm
#> language (EN)
#> collate English_United States.utf8
#> ctype English_United States.utf8
#> tz Asia/Tehran
#> date 2024-05-14
#> pandoc 3.1.1 @ D:/R/RStudio/resources/app/bin/quarto/bin/tools/ (via rmarkdown)
#>
#> ─ Packages ───────────────────────────────────────────────────────────────────
#> package * version date (UTC) lib source
#> backports 1.4.1 2021-12-13 [1] CRAN (R 4.2.0)
#> bbotk 0.7.3 2023-11-13 [1] CRAN (R 4.2.3)
#> cachem 1.0.8 2023-05-01 [1] CRAN (R 4.2.3)
#> callr 3.7.3 2022-11-02 [1] CRAN (R 4.2.3)
#> checkmate 2.3.0 2023-10-25 [1] CRAN (R 4.2.3)
#> class 7.3-22 2023-05-03 [2] CRAN (R 4.3.2)
#> cli 3.6.1 2023-03-23 [1] CRAN (R 4.2.3)
#> clue 0.3-65 2023-09-23 [1] CRAN (R 4.2.3)
#> cluster 2.1.4 2022-08-22 [2] CRAN (R 4.3.2)
#> codetools 0.2-19 2023-02-01 [2] CRAN (R 4.3.2)
#> colorspace 2.1-0 2023-01-23 [1] CRAN (R 4.2.3)
#> crayon 1.5.2 2022-09-29 [1] CRAN (R 4.2.3)
#> data.table 1.14.8 2023-02-17 [1] CRAN (R 4.2.3)
#> DEoptimR 1.1-3 2023-10-07 [1] CRAN (R 4.2.3)
#> devtools 2.4.5 2022-10-11 [1] CRAN (R 4.2.3)
#> dictionar6 0.1.3 2021-09-13 [1] CRAN (R 4.2.3)
#> digest 0.6.33 2023-07-07 [1] CRAN (R 4.2.3)
#> diptest 0.77-0 2023-11-27 [1] CRAN (R 4.2.3)
#> distr6 1.8.4 2023-12-09 [1] Github (xoopR/distr6@1854b22)
#> dplyr * 1.1.4 2023-11-17 [1] CRAN (R 4.2.3)
#> ellipsis 0.3.2 2021-04-29 [1] CRAN (R 4.2.3)
#> evaluate 0.22 2023-09-29 [1] CRAN (R 4.2.3)
#> fansi 1.0.5 2023-10-08 [1] CRAN (R 4.2.3)
#> fastmap 1.1.1 2023-02-24 [1] CRAN (R 4.2.3)
#> flexmix 2.3-19 2023-03-16 [1] CRAN (R 4.2.3)
#> forcats * 1.0.0 2023-01-29 [1] CRAN (R 4.2.3)
#> fpc 2.2-10 2023-01-07 [1] CRAN (R 4.2.3)
#> fs 1.6.3 2023-07-20 [1] CRAN (R 4.2.3)
#> future 1.33.0 2023-07-01 [1] CRAN (R 4.2.3)
#> future.apply 1.11.0 2023-05-21 [1] CRAN (R 4.2.3)
#> generics 0.1.3 2022-07-05 [1] CRAN (R 4.2.3)
#> ggplot2 * 3.4.4 2023-10-12 [1] CRAN (R 4.2.3)
#> globals 0.16.2 2022-11-21 [1] CRAN (R 4.2.2)
#> glue 1.6.2 2022-02-24 [1] CRAN (R 4.2.3)
#> gtable 0.3.4 2023-08-21 [1] CRAN (R 4.2.3)
#> hms 1.1.3 2023-03-21 [1] CRAN (R 4.2.3)
#> htmltools 0.5.6.1 2023-10-06 [1] CRAN (R 4.2.3)
#> htmlwidgets 1.6.2 2023-03-17 [1] CRAN (R 4.2.3)
#> httpuv 1.6.11 2023-05-11 [1] CRAN (R 4.2.3)
#> jsonlite 1.8.7 2023-06-29 [1] CRAN (R 4.2.3)
#> kernlab 0.9-32 2023-01-31 [1] CRAN (R 4.2.2)
#> knitr 1.45 2023-10-30 [1] CRAN (R 4.2.3)
#> later 1.3.1 2023-05-02 [1] CRAN (R 4.2.3)
#> lattice 0.21-9 2023-10-01 [2] CRAN (R 4.3.2)
#> lgr 0.4.4 2022-09-05 [1] CRAN (R 4.2.3)
#> lifecycle 1.0.3 2022-10-07 [1] CRAN (R 4.2.3)
#> listenv 0.9.0 2022-12-16 [1] CRAN (R 4.2.3)
#> lubridate * 1.9.3 2023-09-27 [1] CRAN (R 4.2.3)
#> magrittr 2.0.3 2022-03-30 [1] CRAN (R 4.2.3)
#> MASS 7.3-60 2023-05-04 [1] CRAN (R 4.2.3)
#> Matrix 1.6-4 2023-11-30 [1] CRAN (R 4.2.3)
#> mclust 6.0.1 2023-11-15 [1] CRAN (R 4.2.3)
#> memoise 2.0.1 2021-11-26 [1] CRAN (R 4.2.3)
#> mime 0.12 2021-09-28 [1] CRAN (R 4.2.0)
#> miniUI 0.1.1.1 2018-05-18 [1] CRAN (R 4.2.3)
#> mlr3 * 0.17.0.9000 2023-12-09 [1] Github (mlr-org/mlr3@dc2a983)
#> mlr3cluster 0.1.8 2023-03-12 [1] CRAN (R 4.2.3)
#> mlr3data 0.7.0 2023-06-29 [1] CRAN (R 4.2.3)
#> mlr3extralearners 0.7.1-9000 2023-12-09 [1] Github (mlr-org/mlr3extralearners@7546845)
#> mlr3filters * 0.7.1 2023-02-15 [1] CRAN (R 4.2.3)
#> mlr3fselect 0.11.0 2023-03-02 [1] CRAN (R 4.2.3)
#> mlr3hyperband 0.4.5 2023-03-02 [1] CRAN (R 4.2.3)
#> mlr3learners * 0.5.7 2023-11-21 [1] CRAN (R 4.2.3)
#> mlr3mbo 0.2.1 2023-06-05 [1] CRAN (R 4.2.3)
#> mlr3misc 0.13.0 2023-09-20 [1] CRAN (R 4.2.3)
#> mlr3pipelines 0.5.0-2 2023-12-08 [1] CRAN (R 4.2.3)
#> mlr3proba * 0.5.4 2023-12-09 [1] Github (mlr-org/mlr3proba@083e685)
#> mlr3tuning 0.19.2 2023-11-28 [1] CRAN (R 4.2.3)
#> mlr3tuningspaces 0.4.0 2023-04-20 [1] CRAN (R 4.3.2)
#> mlr3verse * 0.2.8 2023-12-13 [1] Github (mlr-org/mlr3verse@2220087)
#> mlr3viz 0.6.2 2023-11-23 [1] CRAN (R 4.2.3)
#> modeltools 0.2-23 2020-03-05 [1] CRAN (R 4.2.0)
#> munsell 0.5.0 2018-06-12 [1] CRAN (R 4.2.3)
#> nnet 7.3-19 2023-05-03 [2] CRAN (R 4.3.2)
#> ooplah 0.2.0 2022-01-21 [1] CRAN (R 4.2.3)
#> palmerpenguins 0.1.1 2022-08-15 [1] CRAN (R 4.2.3)
#> paradox 0.11.1 2023-03-17 [1] CRAN (R 4.2.3)
#> parallelly 1.36.0 2023-05-26 [1] CRAN (R 4.2.3)
#> param6 0.2.4 2023-12-09 [1] Github (xoopR/param6@0fa3577)
#> pillar 1.9.0 2023-03-22 [1] CRAN (R 4.2.3)
#> pkgbuild 1.4.2 2023-06-26 [1] CRAN (R 4.2.3)
#> pkgconfig 2.0.3 2019-09-22 [1] CRAN (R 4.2.3)
#> pkgload 1.3.3 2023-09-22 [1] CRAN (R 4.2.3)
#> prabclus 2.3-3 2023-10-24 [1] CRAN (R 4.2.3)
#> prettyunits 1.2.0 2023-09-24 [1] CRAN (R 4.2.3)
#> processx 3.8.2 2023-06-30 [1] CRAN (R 4.2.3)
#> profvis 0.3.8 2023-05-02 [1] CRAN (R 4.2.3)
#> promises 1.2.1 2023-08-10 [1] CRAN (R 4.2.3)
#> ps 1.7.5 2023-04-18 [1] CRAN (R 4.2.3)
#> purrr * 1.0.2 2023-08-10 [1] CRAN (R 4.2.3)
#> R.cache 0.16.0 2022-07-21 [1] CRAN (R 4.2.3)
#> R.methodsS3 1.8.2 2022-06-13 [1] CRAN (R 4.2.2)
#> R.oo 1.25.0 2022-06-12 [1] CRAN (R 4.2.2)
#> R.utils 2.12.2 2022-11-11 [1] CRAN (R 4.2.3)
#> R6 2.5.1 2021-08-19 [1] CRAN (R 4.2.3)
#> Rcpp 1.0.11 2023-07-06 [1] CRAN (R 4.2.3)
#> readr * 2.1.4 2023-02-10 [1] CRAN (R 4.2.3)
#> remotes 2.4.2.1 2023-07-18 [1] CRAN (R 4.2.3)
#> reprex * 2.1.0 2024-01-11 [1] CRAN (R 4.3.2)
#> rlang 1.1.1 2023-04-28 [1] CRAN (R 4.2.3)
#> rmarkdown 2.25 2023-09-18 [1] CRAN (R 4.2.3)
#> robustbase 0.99-1 2023-11-29 [1] CRAN (R 4.2.3)
#> rstudioapi 0.15.0 2023-07-07 [1] CRAN (R 4.2.3)
#> scales 1.2.1 2022-08-20 [1] CRAN (R 4.2.3)
#> sessioninfo 1.2.2 2021-12-06 [1] CRAN (R 4.2.3)
#> set6 0.2.6 2023-12-09 [1] Github (xoopR/set6@a901255)
#> shiny 1.7.5 2023-08-12 [1] CRAN (R 4.2.3)
#> spacefillr 0.3.2 2022-10-25 [1] CRAN (R 4.2.3)
#> stringi 1.7.12 2023-01-11 [1] CRAN (R 4.2.2)
#> stringr * 1.5.0 2022-12-02 [1] CRAN (R 4.2.3)
#> styler 1.10.2 2023-08-29 [1] CRAN (R 4.2.3)
#> survival * 3.5-7 2023-08-14 [1] CRAN (R 4.2.3)
#> tibble * 3.2.1 2023-03-20 [1] CRAN (R 4.2.3)
#> tidyr * 1.3.0 2023-01-24 [1] CRAN (R 4.2.3)
#> tidyselect 1.2.0 2022-10-10 [1] CRAN (R 4.2.3)
#> tidyverse * 2.0.0 2023-02-22 [1] CRAN (R 4.2.3)
#> timechange 0.2.0 2023-01-11 [1] CRAN (R 4.2.3)
#> tzdb 0.4.0 2023-05-12 [1] CRAN (R 4.2.3)
#> urlchecker 1.0.1 2021-11-30 [1] CRAN (R 4.2.3)
#> usethis 2.2.2 2023-07-06 [1] CRAN (R 4.2.3)
#> utf8 1.2.3 2023-01-31 [1] CRAN (R 4.2.3)
#> uuid 1.1-1 2023-08-17 [1] CRAN (R 4.2.3)
#> vctrs 0.6.4 2023-10-12 [1] CRAN (R 4.2.3)
#> withr 2.5.1 2023-09-26 [1] CRAN (R 4.2.3)
#> xfun 0.40 2023-08-09 [1] CRAN (R 4.2.3)
#> xgboost * 1.7.6.1 2023-12-06 [1] CRAN (R 4.2.3)
#> xtable 1.8-4 2019-04-21 [1] CRAN (R 4.2.3)
#> yaml 2.3.7 2023-01-23 [1] CRAN (R 4.2.3)
#>
#> [1] D:/R/Libraries
#> [2] C:/Program Files/R/R-4.3.2/library
#>
#> ──────────────────────────────────────────────────────────────────────────────Created on 2024-05-14 with reprex v2.1.0 `
bblodfon
commented
4 months ago Hi @fa1999abdi. Please read the documentation and spend more time understanding what each function you use does via the given examples. The problems you had, had nothing to do with mlr3proba, so it shouldn't be an issue here. Please ask questions like these in stackoverflow.
On your issue: see mlr3fselect::fsi(), you need to select one of mlr_fselectors. The all-in-one solution you can use to do what you want is the mlr3fselect::auto_fselector which has an example. Here is a working example for your code:
library(mlr3verse)
#> Loading required package: mlr3
library(mlr3proba)
task = as_task_surv(x = survival::veteran, time = 'time', event = 'status')
poe = po('encode')
task = poe$train(list(task))[[1]]
set.seed(42)
part = partition(task, ratio = 0.8)
# better "surv.xgboost.cox" to get `distr` predictions as well
xgb = lrn("surv.xgboost.cox", eta = to_tune(1e-4, 1),gamma = to_tune(1e-4, 1),
max_depth = to_tune(1,35), min_child_weight = to_tune(0, 10))
set.seed(456)
at_xgb = auto_tuner(
tuner = tnr("random_search"),
learner = xgb,
resampling = rsmp("cv", folds = 3),
measure = msr("surv.cindex"),
terminator = trm("evals", n_evals = 10)
)
modelxgboost = at_xgb$train(task, part$train)
#> INFO [00:50:46.374] [bbotk] Starting to optimize 4 parameter(s) with '<OptimizerRandomSearch>' and '<TerminatorEvals> [n_evals=10, k=0]'
#> INFO [00:50:46.460] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:46.476] [mlr3] Running benchmark with 3 resampling iterations
#> INFO [00:50:46.516] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/3)
#> INFO [00:50:46.578] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/3)
#> INFO [00:50:46.628] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/3)
#> INFO [00:50:46.874] [mlr3] Finished benchmark
#> INFO [00:50:46.928] [bbotk] Result of batch 1:
#> INFO [00:50:46.933] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [00:50:46.933] [bbotk] 0.6924036 0.06123976 31 8.722243 0.6632974 0 0
#> INFO [00:50:46.933] [bbotk] runtime_learners uhash
#> INFO [00:50:46.933] [bbotk] 0.295 1393e023-c09e-4837-adfd-9c62c489c292
#> INFO [00:50:46.950] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:46.959] [mlr3] Running benchmark with 3 resampling iterations
#> INFO [00:50:46.965] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/3)
#> INFO [00:50:47.016] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/3)
#> INFO [00:50:47.311] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/3)
#> INFO [00:50:47.361] [mlr3] Finished benchmark
#> INFO [00:50:47.400] [bbotk] Result of batch 2:
#> INFO [00:50:47.408] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [00:50:47.408] [bbotk] 0.5470714 0.6802136 29 0.6464914 0.6759085 0 0
#> INFO [00:50:47.408] [bbotk] runtime_learners uhash
#> INFO [00:50:47.408] [bbotk] 0.333 92792d90-4c8b-445a-8403-7f1c5cc09187
#> INFO [00:50:47.426] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:47.434] [mlr3] Running benchmark with 3 resampling iterations
#> INFO [00:50:47.441] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/3)
#> INFO [00:50:47.489] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/3)
#> INFO [00:50:47.539] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/3)
#> INFO [00:50:47.586] [mlr3] Finished benchmark
#> INFO [00:50:47.631] [bbotk] Result of batch 3:
#> INFO [00:50:47.634] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [00:50:47.634] [bbotk] 0.5275122 0.6004684 2 2.587627 0.6602793 0 0
#> INFO [00:50:47.634] [bbotk] runtime_learners uhash
#> INFO [00:50:47.634] [bbotk] 0.089 a597fe3f-0a93-4a36-9287-0620970e887a
#> INFO [00:50:47.651] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:47.659] [mlr3] Running benchmark with 3 resampling iterations
#> INFO [00:50:47.665] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/3)
#> INFO [00:50:47.714] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/3)
#> INFO [00:50:47.779] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/3)
#> INFO [00:50:47.831] [mlr3] Finished benchmark
#> INFO [00:50:47.880] [bbotk] Result of batch 4:
#> INFO [00:50:47.883] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [00:50:47.883] [bbotk] 0.08838163 0.7743576 11 2.235016 0.6677882 0 0
#> INFO [00:50:47.883] [bbotk] runtime_learners uhash
#> INFO [00:50:47.883] [bbotk] 0.107 0c1a3963-702e-408f-b76a-b519ded4dde7
#> INFO [00:50:47.900] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:47.909] [mlr3] Running benchmark with 3 resampling iterations
#> INFO [00:50:47.915] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/3)
#> INFO [00:50:47.972] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/3)
#> INFO [00:50:48.038] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/3)
#> INFO [00:50:48.087] [mlr3] Finished benchmark
#> INFO [00:50:48.132] [bbotk] Result of batch 5:
#> INFO [00:50:48.135] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [00:50:48.135] [bbotk] 0.6383713 0.05744726 8 6.555466 0.688853 0 0
#> INFO [00:50:48.135] [bbotk] runtime_learners uhash
#> INFO [00:50:48.135] [bbotk] 0.112 82a6c452-0ddc-41ae-9aa5-af7bc1d3d352
#> INFO [00:50:48.151] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:48.160] [mlr3] Running benchmark with 3 resampling iterations
#> INFO [00:50:48.166] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/3)
#> INFO [00:50:48.220] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/3)
#> INFO [00:50:48.300] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/3)
#> INFO [00:50:48.357] [mlr3] Finished benchmark
#> INFO [00:50:48.395] [bbotk] Result of batch 6:
#> INFO [00:50:48.398] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [00:50:48.398] [bbotk] 0.125642 0.181675 19 1.169374 0.6812058 0 0
#> INFO [00:50:48.398] [bbotk] runtime_learners uhash
#> INFO [00:50:48.398] [bbotk] 0.131 7ca02bf7-8c0b-4f16-a3cb-288e1da6a5c8
#> INFO [00:50:48.414] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:48.422] [mlr3] Running benchmark with 3 resampling iterations
#> INFO [00:50:48.428] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/3)
#> INFO [00:50:48.480] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/3)
#> INFO [00:50:48.531] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/3)
#> INFO [00:50:48.585] [mlr3] Finished benchmark
#> INFO [00:50:48.622] [bbotk] Result of batch 7:
#> INFO [00:50:48.625] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [00:50:48.625] [bbotk] 0.837198 0.1869602 29 9.887145 0.6682595 0 0
#> INFO [00:50:48.625] [bbotk] runtime_learners uhash
#> INFO [00:50:48.625] [bbotk] 0.097 b8e380be-0baa-433b-8d5b-6ca6956c700e
#> INFO [00:50:48.641] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:48.649] [mlr3] Running benchmark with 3 resampling iterations
#> INFO [00:50:48.656] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/3)
#> INFO [00:50:48.707] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/3)
#> INFO [00:50:48.763] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/3)
#> INFO [00:50:48.812] [mlr3] Finished benchmark
#> INFO [00:50:48.850] [bbotk] Result of batch 8:
#> INFO [00:50:48.853] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [00:50:48.853] [bbotk] 0.07591018 0.8577488 5 5.058832 0.6663479 0 0
#> INFO [00:50:48.853] [bbotk] runtime_learners uhash
#> INFO [00:50:48.853] [bbotk] 0.091 709306e0-7d90-4051-b94e-6b6ede312fb3
#> INFO [00:50:48.870] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:48.879] [mlr3] Running benchmark with 3 resampling iterations
#> INFO [00:50:48.885] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/3)
#> INFO [00:50:48.935] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/3)
#> INFO [00:50:49.043] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/3)
#> INFO [00:50:49.106] [mlr3] Finished benchmark
#> INFO [00:50:49.146] [bbotk] Result of batch 9:
#> INFO [00:50:49.149] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [00:50:49.149] [bbotk] 0.2607173 0.6375698 29 9.87441 0.6691351 0 0
#> INFO [00:50:49.149] [bbotk] runtime_learners uhash
#> INFO [00:50:49.149] [bbotk] 0.163 a4c69de7-a3c2-44bb-8dab-cd6a882b163f
#> INFO [00:50:49.166] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:49.175] [mlr3] Running benchmark with 3 resampling iterations
#> INFO [00:50:49.181] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/3)
#> INFO [00:50:49.240] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/3)
#> INFO [00:50:49.503] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/3)
#> INFO [00:50:49.551] [mlr3] Finished benchmark
#> INFO [00:50:49.591] [bbotk] Result of batch 10:
#> INFO [00:50:49.593] [bbotk] eta gamma max_depth min_child_weight surv.cindex warnings errors
#> INFO [00:50:49.593] [bbotk] 0.8892446 0.334645 5 3.901119 0.6920791 0 0
#> INFO [00:50:49.593] [bbotk] runtime_learners uhash
#> INFO [00:50:49.593] [bbotk] 0.312 c2ab1381-e30e-4ea6-837d-880c905f8ac2
#> INFO [00:50:49.623] [bbotk] Finished optimizing after 10 evaluation(s)
#> INFO [00:50:49.624] [bbotk] Result:
#> INFO [00:50:49.626] [bbotk] eta gamma max_depth min_child_weight learner_param_vals x_domain
#> INFO [00:50:49.626] [bbotk] <num> <num> <int> <num> <list> <list>
#> INFO [00:50:49.626] [bbotk] 0.8892446 0.334645 5 3.901119 <list[8]> <list[4]>
#> INFO [00:50:49.626] [bbotk] surv.cindex
#> INFO [00:50:49.626] [bbotk] <num>
#> INFO [00:50:49.626] [bbotk] 0.6920791
learner = modelxgboost$learner$clone() # get the tuned learner
learner
#> <LearnerSurvXgboostCox:surv.xgboost.cox>: Extreme Gradient Boosting Cox
#> * Model: list
#> * Parameters: early_stopping_set=none, eta=0.8892, gamma=0.3346,
#> max_depth=5, min_child_weight=3.901, nrounds=1, nthread=1, verbose=0
#> * Packages: mlr3, mlr3proba, mlr3extralearners, xgboost
#> * Predict Types: [crank], distr, lp
#> * Feature Types: integer, numeric
#> * Properties: importance, missings, weights
# Recursive Feature Elimination with Cross Validation
instance = fsi(
task = task$clone()$filter(part$train), # How to use train data(part$train)?
learner = learner, # How can I use tuned Lerner in the previous step to Select the variable?
resampling = rsmp("cv", folds = 6),
measures = msr("surv.cindex"),
terminator = trm("none")
)
# optimizer$optimize(instance) # Why does it give an error?
#> Error in eval(expr, envir, enclos): object 'optimizer' not found
# See `mlr3fselect::fsi`, you need to select one of `mlr_fselectors`
# And since you want RFE it has to be that:
fselector = fs("rfe", n_features = 2, feature_fraction = 0.8)
fselector$optimize(instance)
#> INFO [00:50:50.161] [bbotk] Starting to optimize 9 parameter(s) with '<FSelectorRFE>' and '<TerminatorNone>'
#> INFO [00:50:50.163] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:50.172] [mlr3] Running benchmark with 6 resampling iterations
#> INFO [00:50:50.179] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/6)
#> INFO [00:50:50.220] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/6)
#> INFO [00:50:50.261] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/6)
#> INFO [00:50:50.539] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 4/6)
#> INFO [00:50:50.583] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 5/6)
#> INFO [00:50:50.626] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 6/6)
#> INFO [00:50:50.667] [mlr3] Finished benchmark
#> INFO [00:50:50.699] [bbotk] Result of batch 1:
#> INFO [00:50:50.703] [bbotk] age diagtime karno prior trt celltype.squamous celltype.smallcell
#> INFO [00:50:50.703] [bbotk] TRUE TRUE TRUE TRUE TRUE TRUE TRUE
#> INFO [00:50:50.703] [bbotk] celltype.adeno celltype.large surv.cindex warnings errors runtime_learners
#> INFO [00:50:50.703] [bbotk] TRUE TRUE 0.7046958 0 0 0.419
#> INFO [00:50:50.703] [bbotk] uhash
#> INFO [00:50:50.703] [bbotk] a4ff7734-9d05-4cd9-ae0a-8be79807e216
#> INFO [00:50:50.784] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:50.794] [mlr3] Running benchmark with 6 resampling iterations
#> INFO [00:50:50.801] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/6)
#> INFO [00:50:50.844] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/6)
#> INFO [00:50:50.886] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/6)
#> INFO [00:50:50.931] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 4/6)
#> INFO [00:50:50.974] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 5/6)
#> INFO [00:50:51.027] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 6/6)
#> INFO [00:50:51.068] [mlr3] Finished benchmark
#> INFO [00:50:51.103] [bbotk] Result of batch 2:
#> INFO [00:50:51.107] [bbotk] age diagtime karno prior trt celltype.squamous celltype.smallcell
#> INFO [00:50:51.107] [bbotk] TRUE TRUE TRUE FALSE TRUE TRUE TRUE
#> INFO [00:50:51.107] [bbotk] celltype.adeno celltype.large surv.cindex warnings errors runtime_learners
#> INFO [00:50:51.107] [bbotk] FALSE TRUE 0.6971201 0 0 0.194
#> INFO [00:50:51.107] [bbotk] uhash
#> INFO [00:50:51.107] [bbotk] 2fda12ab-a0b1-4035-afd8-dba2e07b9589
#> INFO [00:50:51.186] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:51.194] [mlr3] Running benchmark with 6 resampling iterations
#> INFO [00:50:51.200] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/6)
#> INFO [00:50:51.747] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/6)
#> INFO [00:50:52.038] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/6)
#> INFO [00:50:52.082] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 4/6)
#> INFO [00:50:52.379] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 5/6)
#> INFO [00:50:52.429] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 6/6)
#> INFO [00:50:52.469] [mlr3] Finished benchmark
#> INFO [00:50:52.505] [bbotk] Result of batch 3:
#> INFO [00:50:52.508] [bbotk] age diagtime karno prior trt celltype.squamous celltype.smallcell
#> INFO [00:50:52.508] [bbotk] TRUE TRUE TRUE FALSE FALSE TRUE TRUE
#> INFO [00:50:52.508] [bbotk] celltype.adeno celltype.large surv.cindex warnings errors runtime_learners
#> INFO [00:50:52.508] [bbotk] FALSE FALSE 0.7205039 0 0 0.943
#> INFO [00:50:52.508] [bbotk] uhash
#> INFO [00:50:52.508] [bbotk] 061330b4-861a-4390-b3ae-af7ea26620d9
#> INFO [00:50:52.585] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:52.593] [mlr3] Running benchmark with 6 resampling iterations
#> INFO [00:50:52.599] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/6)
#> INFO [00:50:52.641] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/6)
#> INFO [00:50:52.694] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/6)
#> INFO [00:50:52.744] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 4/6)
#> INFO [00:50:52.785] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 5/6)
#> INFO [00:50:52.828] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 6/6)
#> INFO [00:50:52.868] [mlr3] Finished benchmark
#> INFO [00:50:52.903] [bbotk] Result of batch 4:
#> INFO [00:50:52.906] [bbotk] age diagtime karno prior trt celltype.squamous celltype.smallcell
#> INFO [00:50:52.906] [bbotk] TRUE TRUE TRUE FALSE FALSE FALSE TRUE
#> INFO [00:50:52.906] [bbotk] celltype.adeno celltype.large surv.cindex warnings errors runtime_learners
#> INFO [00:50:52.906] [bbotk] FALSE FALSE 0.6968156 0 0 0.205
#> INFO [00:50:52.906] [bbotk] uhash
#> INFO [00:50:52.906] [bbotk] 87c9810b-6109-4cd4-99af-b88bc91f4a6c
#> INFO [00:50:52.987] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:52.995] [mlr3] Running benchmark with 6 resampling iterations
#> INFO [00:50:53.001] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/6)
#> INFO [00:50:53.287] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/6)
#> INFO [00:50:53.330] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/6)
#> INFO [00:50:53.642] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 4/6)
#> INFO [00:50:53.687] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 5/6)
#> INFO [00:50:53.921] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 6/6)
#> INFO [00:50:53.980] [mlr3] Finished benchmark
#> INFO [00:50:54.017] [bbotk] Result of batch 5:
#> INFO [00:50:54.021] [bbotk] age diagtime karno prior trt celltype.squamous celltype.smallcell
#> INFO [00:50:54.021] [bbotk] TRUE TRUE TRUE FALSE FALSE FALSE FALSE
#> INFO [00:50:54.021] [bbotk] celltype.adeno celltype.large surv.cindex warnings errors runtime_learners
#> INFO [00:50:54.021] [bbotk] FALSE FALSE 0.6935666 0 0 0.902
#> INFO [00:50:54.021] [bbotk] uhash
#> INFO [00:50:54.021] [bbotk] 0226dae0-acae-4e9d-bd95-07bea64fd64f
#> INFO [00:50:54.103] [bbotk] Evaluating 1 configuration(s)
#> INFO [00:50:54.111] [mlr3] Running benchmark with 6 resampling iterations
#> INFO [00:50:54.118] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 1/6)
#> INFO [00:50:54.200] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 2/6)
#> INFO [00:50:54.244] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 3/6)
#> INFO [00:50:54.287] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 4/6)
#> INFO [00:50:54.339] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 5/6)
#> INFO [00:50:54.383] [mlr3] Applying learner 'surv.xgboost.cox' on task 'survival::veteran' (iter 6/6)
#> INFO [00:50:54.424] [mlr3] Finished benchmark
#> INFO [00:50:54.467] [bbotk] Result of batch 6:
#> INFO [00:50:54.470] [bbotk] age diagtime karno prior trt celltype.squamous celltype.smallcell
#> INFO [00:50:54.470] [bbotk] TRUE FALSE TRUE FALSE FALSE FALSE FALSE
#> INFO [00:50:54.470] [bbotk] celltype.adeno celltype.large surv.cindex warnings errors runtime_learners
#> INFO [00:50:54.470] [bbotk] FALSE FALSE 0.7012597 0 0 0.239
#> INFO [00:50:54.470] [bbotk] uhash
#> INFO [00:50:54.470] [bbotk] 8c1472ee-f3c3-4776-be7b-548921dd3be4
#> INFO [00:50:54.557] [bbotk] Finished optimizing after 6 evaluation(s)
#> INFO [00:50:54.558] [bbotk] Result:
#> INFO [00:50:54.561] [bbotk] age diagtime karno prior trt celltype.squamous celltype.smallcell
#> INFO [00:50:54.561] [bbotk] <lgcl> <lgcl> <lgcl> <lgcl> <lgcl> <lgcl> <lgcl>
#> INFO [00:50:54.561] [bbotk] TRUE TRUE TRUE FALSE FALSE TRUE TRUE
#> INFO [00:50:54.561] [bbotk] celltype.adeno celltype.large importance
#> INFO [00:50:54.561] [bbotk] <lgcl> <lgcl> <list>
#> INFO [00:50:54.561] [bbotk] FALSE FALSE 5.000000,3.666667,2.833333,1.833333,1.666667
#> INFO [00:50:54.561] [bbotk] features n_features surv.cindex
#> INFO [00:50:54.561] [bbotk] <list> <int> <num>
#> INFO [00:50:54.561] [bbotk] age,diagtime,karno,celltype.squamous,celltype.smallcell 5 0.7205039
#> age diagtime karno prior trt celltype.squamous celltype.smallcell
#> <lgcl> <lgcl> <lgcl> <lgcl> <lgcl> <lgcl> <lgcl>
#> 1: TRUE TRUE TRUE FALSE FALSE TRUE TRUE
#> celltype.adeno celltype.large importance
#> <lgcl> <lgcl> <list>
#> 1: FALSE FALSE 5.000000,3.666667,2.833333,1.833333,1.666667
#> features n_features
#> <list> <int>
#> 1: age,diagtime,karno,celltype.squamous,celltype.smallcell 5
#> surv.cindex
#> <num>
#> 1: 0.7205039
as.data.table(instance$archive)
#> age diagtime karno prior trt celltype.squamous celltype.smallcell
#> <lgcl> <lgcl> <lgcl> <lgcl> <lgcl> <lgcl> <lgcl>
#> 1: TRUE TRUE TRUE TRUE TRUE TRUE TRUE
#> 2: TRUE TRUE TRUE FALSE TRUE TRUE TRUE
#> 3: TRUE TRUE TRUE FALSE FALSE TRUE TRUE
#> 4: TRUE TRUE TRUE FALSE FALSE FALSE TRUE
#> 5: TRUE TRUE TRUE FALSE FALSE FALSE FALSE
#> 6: TRUE FALSE TRUE FALSE FALSE FALSE FALSE
#> celltype.adeno celltype.large surv.cindex runtime_learners
#> <lgcl> <lgcl> <num> <num>
#> 1: TRUE TRUE 0.7046958 0.419
#> 2: FALSE TRUE 0.6971201 0.194
#> 3: FALSE FALSE 0.7205039 0.943
#> 4: FALSE FALSE 0.6968156 0.205
#> 5: FALSE FALSE 0.6935666 0.902
#> 6: FALSE FALSE 0.7012597 0.239
#> timestamp batch_nr warnings errors
#> <POSc> <int> <int> <int>
#> 1: 2024-05-16 00:50:50 1 0 0
#> 2: 2024-05-16 00:50:51 2 0 0
#> 3: 2024-05-16 00:50:52 3 0 0
#> 4: 2024-05-16 00:50:52 4 0 0
#> 5: 2024-05-16 00:50:54 5 0 0
#> 6: 2024-05-16 00:50:54 6 0 0
#> importance
#> <list>
#> 1: 9.000000,7.666667,6.166667,5.333333,4.833333,4.166667,...
#> 2: 7.000000,5.666667,4.166667,3.833333,2.833333,2.833333,...
#> 3: 5.000000,3.666667,2.833333,1.833333,1.666667
#> 4: 4.000000,2.666667,2.000000,1.333333
#> 5: 3,2,1
#> 6: 2,1
#> features n_features
#> <list> <list>
#> 1: age,diagtime,karno,prior,trt,celltype.squamous,... 9
#> 2: age,diagtime,karno,trt,celltype.squamous,celltype.smallcell,... 7
#> 3: age,diagtime,karno,celltype.squamous,celltype.smallcell 5
#> 4: age,diagtime,karno,celltype.smallcell 4
#> 5: age,diagtime,karno 3
#> 6: age,karno 2
#> resample_result
#> <list>
#> 1: <ResampleResult>
#> 2: <ResampleResult>
#> 3: <ResampleResult>
#> 4: <ResampleResult>
#> 5: <ResampleResult>
#> 6: <ResampleResult>
# I also want to subset the task to the optimal feature set
# and again train the learner.
task$select(instance$result_feature_set)
learner$train(task, part$train)
# And finally, I use the trained model to predict the survival time on test data,
# which will encounter an error.
p = learner$predict(task, part$test)
p
#> <PredictionSurv> for 28 observations:
#> row_ids time status crank lp distr
#> 1 72 TRUE -0.458429651 -0.458429651 <list[1]>
#> 7 82 TRUE -0.255330144 -0.255330144 <list[1]>
#> 13 144 TRUE 1.800571186 1.800571186 <list[1]>
#> ---
#> 135 231 TRUE -0.644105278 -0.644105278 <list[1]>
#> 21 123 FALSE 1.800571186 1.800571186 <list[1]>
#> 22 97 FALSE 0.005314152 0.005314152 <list[1]>Created on 2024-05-16 with reprex v2.0.2
bblodfon
commented
4 months ago Also, what you are doing might not be optimal, since you first tune the learner using all features (so hyperparameters get fixed) and then you do RFE with the same hyperparameters for each different subset, whereas it might be the case that each different subset might need different optimal hyperparameters. If you wanna do this, which means way heavier computation as well, it would mean nested resampling, see this post.
fa1999abdi
commented
4 months ago tanks john
also, I asked in issue #363 about getting the survival time prediction, but it doesn't seem like the response column is the real survival time prediction,
Why are survival times so close for individuals with different times and statuses? (Why row_ids 2 and 3 have exactly the same response with different times and statuses? What does that mean?)
bblodfon
commented
4 months ago It's the cumulative hazard (method) which as you say doesn't correspond to a true survival time. In this case this happens because its a crank composition and we added the response as an extra (but didn't think too much about the actual values). The cum_haz uses the distr prediction and the computation is very simple, see ?survivalmodels::surv_to_risk(). Also done in RSFs. In your case the two distr must be the same to get the same response.
In general, I am aware of this issue and wanted to change the default response in the composition to be the RMST (restricted mean survival time) which a very nice method for expected survival time, backed by literature. Will ping you when it is done in the Discussion.
fa1999abdi
commented
4 months ago With these explanations, How can I access (predictions$survival_probabilities) to Plotting survival curves for test set?
bblodfon
commented
4 months ago fa1999abdi
commented
3 months ago Thanks so much for your help.
"I'm planning to do a machine learning study. I want to optimize the hyperparameters, train and benchmark the model using the train set, and then measure the accuracy of the model using the test set, and finally analyze and I want to do time-dependent ROC analysis and get the AUC Roc. I have read the MLR 3 book, but I'm having trouble understanding how to code this particular section. Can you help me complete this part of my study?"