R Under development (unstable) (2026-01-07 r89288 ucrt) -- "Unsuffered Consequences" Copyright (C) 2026 The R Foundation for Statistical Computing Platform: x86_64-w64-mingw32/x64 R is free software and comes with ABSOLUTELY NO WARRANTY. You are welcome to redistribute it under certain conditions. Type 'license()' or 'licence()' for distribution details. R is a collaborative project with many contributors. Type 'contributors()' for more information and 'citation()' on how to cite R or R packages in publications. Type 'demo()' for some demos, 'help()' for on-line help, or 'help.start()' for an HTML browser interface to help. Type 'q()' to quit R. > Sys.setenv("R_TESTS" = "") > library(testthat) > library(personalized) Loading required package: glmnet Loading required package: Matrix Loaded glmnet 4.1-10 Loading required package: mgcv Loading required package: nlme This is mgcv 1.9-4. For overview type '?mgcv'. Loading required package: ggplot2 Loading required package: plotly Attaching package: 'plotly' The following object is masked from 'package:ggplot2': last_plot The following object is masked from 'package:stats': filter The following object is masked from 'package:graphics': layout > > test_check("personalized") family: gaussian loss: sq_loss_lasso method: weighting cutpoint: 0 propensity function: propensity.func benefit score: f(x), Trt recom = 1*I(f(x)>c)+0*I(f(x)<=c) where c is 'cutpoint' Average Outcomes: Recommended 0 Recommended 1 Received 0 8.9342 (n = 24) -9.2993 (n = 17) Received 1 -5.888 (n = 29) 7.2975 (n = 30) Treatment effects conditional on subgroups: Est of E[Y|T=0,Recom=0]-E[Y|T=/=0,Recom=0] 14.8221 (n = 53) Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 16.5969 (n = 47) NOTE: The above average outcomes are biased estimates of the expected outcomes conditional on subgroups. Use 'validate.subgroup()' to obtain unbiased estimates. --------------------------------------------------- Benefit score quantiles (f(X) for 1 vs 0): 0% 25% 50% 75% 100% -21.9806 -6.3018 -0.6354 6.1213 25.5801 --------------------------------------------------- Summary of individual treatment effects: E[Y|T=1, X] - E[Y|T=0, X] Min. 1st Qu. Median Mean 3rd Qu. Max. -43.9612 -12.6036 -1.2708 0.1229 12.2425 51.1602 family: gaussian loss: sq_loss_lasso method: weighting cutpoint: 0 propensity function: propensity.func benefit score: f(x), Trt recom = 1*I(f(x)=c) where c is 'cutpoint' Average Outcomes: Recommended 0 Recommended 1 Received 0 -9.1702 (n = 17) 8.2552 (n = 24) Received 1 7.1945 (n = 31) -5.9147 (n = 28) Treatment effects conditional on subgroups: Est of E[Y|T=0,Recom=0]-E[Y|T=/=0,Recom=0] -16.3647 (n = 48) Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] -14.1699 (n = 52) NOTE: The above average outcomes are biased estimates of the expected outcomes conditional on subgroups. Use 'validate.subgroup()' to obtain unbiased estimates. --------------------------------------------------- Benefit score quantiles (f(X) for 1 vs 0): 0% 25% 50% 75% 100% -21.1598 -6.0406 -0.5452 6.0647 25.1260 --------------------------------------------------- Summary of individual treatment effects: E[Y|T=1, X] - E[Y|T=0, X] Min. 1st Qu. Median Mean 3rd Qu. Max. -42.3197 -12.0812 -1.0904 0.3885 12.1294 50.2519 Multiple eval metrics are present. Will use test_wtd_rmse for early stopping. Will train until test_wtd_rmse hasn't improved in 50 rounds. [1] train-wtd_rmse:9.256776±0.391546 test-wtd_rmse:9.618186±0.598393 [2] train-wtd_rmse:8.903396±0.472580 test-wtd_rmse:9.602023±0.667385 [3] train-wtd_rmse:8.553518±0.515614 test-wtd_rmse:9.422115±0.767464 [4] train-wtd_rmse:8.271810±0.559907 test-wtd_rmse:9.433930±0.799954 [5] train-wtd_rmse:8.050969±0.446031 test-wtd_rmse:9.123869±0.773622 Summary of individual treatment effects: E[Y|T=1, X] - E[Y|T=0, X] Min. 1st Qu. Median Mean 3rd Qu. Max. -43.913 -10.825 -1.574 -1.009 8.205 38.722 Summary of individual treatment effects: E[Y|T=1, X] - E[Y|T=0, X] Min. 1st Qu. Median Mean 3rd Qu. Max. -34.390 -5.795 1.886 2.259 10.456 37.506 Summary of individual treatment effects: E[Y|T=1, X] - E[Y|T=0, X] Min. 1st Qu. Median Mean 3rd Qu. Max. -34.390 -5.795 1.886 2.259 10.456 37.506 Summary of individual treatment effects: E[Y|T=1, X] / E[Y|T=0, X] Note: for survival outcomes, the above ratio is E[g(Y)|T=1, X] / E[g(Y)|T=0, X], where g() is a monotone increasing function of Y, the survival time Min. 1st Qu. Median Mean 3rd Qu. Max. 0.2452 0.8272 1.1267 1.2324 1.5091 3.9228 family: binomial loss: logistic_loss_lasso method: weighting validation method: training_test_replication cutpoint: 0 replications: 3 benefit score: f(x), Trt recom = 1*I(f(x)>c)+0*I(f(x)<=c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 0 Recommended 1 Received 0 1 (SE = 0, n = 3.6667) 0.0424 (SE = 0.0735, n = 5.6667) Received 1 0.0331 (SE = 0.0573, n = 7.3333) 0.9339 (SE = 0.1145, n = 8.3333) Treatment effects conditional on subgroups: Est of E[Y|T=0,Recom=0]-E[Y|T=/=0,Recom=0] 0.9669 (SE = 0.0573, n = 11) Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 0.8915 (SE = 0.188, n = 14) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: 0.9067 (SE = 0.1213) family: gaussian loss: sq_loss_lasso method: weighting validation method: training_test_replication cutpoint: 0 replications: 3 benefit score: f(x), Trt recom = 1*I(f(x)=c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 0 Recommended 1 Received 0 -27.101 (SE = 5.5353, n = 2.6667) 21.9228 (SE = 4.1631, n = 7.6667) Received 1 12.1774 (SE = 1.8509, n = 8.6667) -12.9406 (SE = 8.3275, n = 6) Treatment effects conditional on subgroups: Est of E[Y|T=0,Recom=0]-E[Y|T=/=0,Recom=0] -39.2784 (SE = 5.2129, n = 11.3333) Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] -34.8635 (SE = 12.2911, n = 13.6667) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: -38.4093 (SE = 4.2242) family: binomial loss: logistic_loss_lasso method: weighting validation method: training_test_replication cutpoint: Quant_67 replications: 3 benefit score: f(x), Trt recom = 1*I(f(x)>c)+0*I(f(x)<=c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 0 Recommended 1 Received 0 0.8827 (SE = 0.1035, n = 4.6667) 0 (SE = 0, n = 4.6667) Received 1 0.3673 (SE = 0.0878, n = 12.3333) 1 (SE = 0, n = 3.3333) Treatment effects conditional on subgroups: Est of E[Y|T=0,Recom=0]-E[Y|T=/=0,Recom=0] 0.5154 (SE = 0.1634, n = 17) Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 1 (SE = 0, n = 8) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: 0.704 (SE = 0.1303) <===============================================> family: binomial loss: logistic_loss_lasso method: weighting validation method: training_test_replication cutpoint: Quant_83 replications: 3 benefit score: f(x), Trt recom = 1*I(f(x)>c)+0*I(f(x)<=c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 0 Recommended 1 Received 0 0.5684 (SE = 0.1514, n = 6.3333) 0 (SE = 0, n = 3) Received 1 0.4308 (SE = 0.065, n = 13.6667) 1 (SE = 0, n = 2) Treatment effects conditional on subgroups: Est of E[Y|T=0,Recom=0]-E[Y|T=/=0,Recom=0] 0.1377 (SE = 0.1768, n = 20) Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 1 (SE = 0, n = 5) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: 0.3255 (SE = 0.1659) family: binomial loss: logistic_loss_lasso method: weighting validation method: training_test_replication cutpoint: Quant_67 replications: 3 benefit score: f(x), Trt recom = 1*I(f(x)>c)+0*I(f(x)<=c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 0 Recommended 1 Received 0 0.8827 (SE = 0.1035, 18.6667%) 0 (SE = 0, 18.6667%) Received 1 0.3673 (SE = 0.0878, 49.3333%) 1 (SE = 0, 13.3333%) Treatment effects conditional on subgroups: Est of E[Y|T=0,Recom=0]-E[Y|T=/=0,Recom=0] 0.5154 (SE = 0.1634, 68%) Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 1 (SE = 0, 32%) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: 0.704 (SE = 0.1303) <===============================================> family: binomial loss: logistic_loss_lasso method: weighting validation method: training_test_replication cutpoint: Quant_83 replications: 3 benefit score: f(x), Trt recom = 1*I(f(x)>c)+0*I(f(x)<=c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 0 Recommended 1 Received 0 0.5684 (SE = 0.1514, 25.3333%) 0 (SE = 0, 12%) Received 1 0.4308 (SE = 0.065, 54.6667%) 1 (SE = 0, 8%) Treatment effects conditional on subgroups: Est of E[Y|T=0,Recom=0]-E[Y|T=/=0,Recom=0] 0.1377 (SE = 0.1768, 80%) Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 1 (SE = 0, 20%) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: 0.3255 (SE = 0.1659) family: cox loss: cox_loss_lasso method: weighting validation method: training_test_replication cutpoint: 0 replications: 3 benefit score: f(x), Trt recom = 1*I(f(x)>c)+0*I(f(x)<=c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 0 Recommended 1 Received 0 24.4288 (SE = 30.2323, n = 4) 0 (SE = 0, n = 4.6667) Received 1 0 (SE = 0, n = 7.6667) 1.1995 (SE = 0.7975, n = 8.6667) Treatment effects conditional on subgroups: Est of E[Y|T=0,Recom=0]-E[Y|T=/=0,Recom=0] 24.4288 (SE = 30.2323, n = 11.6667) Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 1.1995 (SE = 0.7975, n = 13.3333) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: 14.7452 (SE = 12.9127) family: cox loss: cox_loss_lasso method: weighting validation method: boot_bias_correction cutpoint: 0 replications: 3 benefit score: f(x), Trt recom = 1*I(f(x)>c)+0*I(f(x)<=c) where c is 'cutpoint' Average Bootstrap Bias-Corrected Outcomes: Recommended 0 Recommended 1 Received 0 31.5529 (SE = 16.8759, n = 17.3333) 0 (SE = 0, n = 25.6667) Received 1 0 (SE = 0, n = 31.6667) 1.8959 (SE = 0.891, n = 25.3333) Treatment effects conditional on subgroups: Est of E[Y|T=0,Recom=0]-E[Y|T=/=0,Recom=0] 31.5529 (SE = 16.8759, n = 49) Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 1.8959 (SE = 0.891, n = 51) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: 34.3274 (SE = 11.4573) family: gaussian loss: sq_loss_lasso method: weighting cutpoint: 0 propensity function: propensity.func benefit score: f_2(x): 2 vs 1, f_3(x): 3 vs 1 f_1(x): 0 maxval = max(f_2(x), f_3(x)) which.max(maxval) = The trt level which maximizes maxval Trt recom = which.max(maxval)*I(maxval > c) + 1*I(maxval <= c) where c is 'cutpoint' Average Outcomes: Recommended 1 Recommended 2 Recommended 3 Received 1 19.7513 (n = 4) 15.9236 (n = 28) 23.9965 (n = 1) Received 2 -13.9114 (n = 2) 31.9898 (n = 6) -15.5207 (n = 34) Received 3 -28.2337 (n = 5) -41.1735 (n = 6) 29.1472 (n = 14) Treatment effects conditional on subgroups: Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 41.5168 (n = 11) Est of E[Y|T=2,Recom=2]-E[Y|T=/=2,Recom=2] 30.0126 (n = 40) Est of E[Y|T=3,Recom=3]-E[Y|T=/=3,Recom=3] 41.6508 (n = 49) NOTE: The above average outcomes are biased estimates of the expected outcomes conditional on subgroups. Use 'validate.subgroup()' to obtain unbiased estimates. --------------------------------------------------- Benefit score 1 quantiles (f(X) for 2 vs 1): 0% 25% 50% 75% 100% -52.419 -18.669 -1.927 13.652 61.772 Benefit score 2 quantiles (f(X) for 3 vs 1): 0% 25% 50% 75% 100% -103.787 -30.817 3.594 34.699 105.366 --------------------------------------------------- Summary of individual treatment effects: E[Y|T=trt, X] - E[Y|T=1, X] where 'trt' is 2 and 3 2-vs-1 3-vs-1 Min. :-104.839 Min. :-207.573 1st Qu.: -37.338 1st Qu.: -61.633 Median : -3.855 Median : 7.188 Mean : -1.162 Mean : 2.327 3rd Qu.: 27.303 3rd Qu.: 69.399 Max. : 123.544 Max. : 210.733 family: gaussian loss: sq_loss_lasso method: weighting validation method: training_test_replication cutpoint: 0 replications: 2 benefit score: f_2(x): 2 vs 1, f_3(x): 3 vs 1 f_1(x): 0 maxval = max(f_2(x), f_3(x)) which.max(maxval) = The trt level which maximizes maxval Trt recom = which.max(maxval)*I(maxval > c) + 1*I(maxval <= c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 1 Recommended 2 Received 1 10.9365 (SE = 18.2443, n = 2) 15.0693 (SE = 11.5189, n = 6.5) Received 2 NaN (SE = NA, n = 0) 18.6765 (SE = NA, n = 1) Received 3 -25.872 (SE = 4.9554, n = 1.5) 17.1442 (SE = NA, n = 0.5) Recommended 3 Received 1 23.9965 (SE = NA, n = 0.5) Received 2 -18.3166 (SE = 3.866, n = 9.5) Received 3 45.3389 (SE = 5.7727, n = 3.5) Treatment effects conditional on subgroups: Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 36.8085 (SE = 23.1997, n = 3.5) Est of E[Y|T=2,Recom=2]-E[Y|T=/=2,Recom=2] 11.7522 (SE = NA, n = 8) Est of E[Y|T=3,Recom=3]-E[Y|T=/=3,Recom=3] 58.4252 (SE = 9.3034, n = 13.5) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: 39.3959 (SE = 13.392) family: gaussian loss: sq_loss_lasso method: weighting validation method: training_test_replication cutpoint: Quant_33 replications: 2 benefit score: f_2(x): 2 vs 1, f_3(x): 3 vs 1 f_1(x): 0 maxval = max(f_2(x), f_3(x)) which.max(maxval) = The trt level which maximizes maxval Trt recom = which.max(maxval)*I(maxval > c) + 1*I(maxval <= c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 1 Recommended 2 Received 1 NaN (SE = NA, n = 0) 16.7838 (SE = 1.8647, n = 8.5) Received 2 NaN (SE = NA, n = 0) 18.6765 (SE = NA, n = 1) Received 3 -44.1777 (SE = NA, n = 0.5) 10.5345 (SE = NA, n = 1) Recommended 3 Received 1 23.9965 (SE = NA, n = 0.5) Received 2 -18.3166 (SE = 3.866, n = 9.5) Received 3 43.4349 (SE = 3.0802, n = 4) Treatment effects conditional on subgroups: Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] NaN (SE = NA, n = 0.5) Est of E[Y|T=2,Recom=2]-E[Y|T=/=2,Recom=2] 3.2112 (SE = NA, n = 10.5) Est of E[Y|T=3,Recom=3]-E[Y|T=/=3,Recom=3] 56.5212 (SE = 6.6108, n = 14) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: 39.5889 (SE = 4.6787) <===============================================> family: gaussian loss: sq_loss_lasso method: weighting validation method: training_test_replication cutpoint: Quant_67 replications: 2 benefit score: f_2(x): 2 vs 1, f_3(x): 3 vs 1 f_1(x): 0 maxval = max(f_2(x), f_3(x)) which.max(maxval) = The trt level which maximizes maxval Trt recom = which.max(maxval)*I(maxval > c) + 1*I(maxval <= c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 1 Recommended 2 Received 1 17.8705 (SE = 2.6598, n = 4) 17.1043 (SE = 11.472, n = 5) Received 2 18.6765 (SE = NA, n = 1) NaN (SE = NA, n = 0) Received 3 -12.0197 (SE = 24.5456, n = 3.5) NaN (SE = NA, n = 0) Recommended 3 Received 1 NaN (SE = NA, n = 0) Received 2 -18.3166 (SE = 3.866, n = 9.5) Received 3 52.8108 (SE = 4.7941, n = 2) Treatment effects conditional on subgroups: Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] 19.2468 (SE = 12.1534, n = 8.5) Est of E[Y|T=2,Recom=2]-E[Y|T=/=2,Recom=2] NaN (SE = NA, n = 5) Est of E[Y|T=3,Recom=3]-E[Y|T=/=3,Recom=3] 71.1273 (SE = 8.6602, n = 11.5) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: 42.2847 (SE = 5.1894) family: gaussian loss: sq_loss_lasso method: weighting cutpoint: 0 propensity function: propensity.func benefit score: f_2(x): 2 vs 1, f_3(x): 3 vs 1 f_1(x): 0 minval = min(f_2(x), f_3(x)) which.min(minval) = The trt level which mininizes minval Trt recom = which.min(minval)*I(minval < c) + 1*I(minval >= c) where c is 'cutpoint' Average Outcomes: Recommended 1 Recommended 2 Recommended 3 Received 1 -12.4319 (n = 3) 23.9965 (n = 1) 20.0737 (n = 29) Received 2 16.5515 (n = 8) -23.5188 (n = 28) 24.3617 (n = 6) Received 3 41.8225 (n = 2) 18.1545 (n = 14) -39.4779 (n = 9) Treatment effects conditional on subgroups: Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] -44.6999 (n = 13) Est of E[Y|T=2,Recom=2]-E[Y|T=/=2,Recom=2] -42.1553 (n = 43) Est of E[Y|T=3,Recom=3]-E[Y|T=/=3,Recom=3] -61.4123 (n = 44) NOTE: The above average outcomes are biased estimates of the expected outcomes conditional on subgroups. Use 'validate.subgroup()' to obtain unbiased estimates. --------------------------------------------------- Benefit score 1 quantiles (f(X) for 2 vs 1): 0% 25% 50% 75% 100% -52.058 -18.664 -2.026 13.390 61.143 Benefit score 2 quantiles (f(X) for 3 vs 1): 0% 25% 50% 75% 100% -103.661 -30.787 3.412 34.372 104.861 --------------------------------------------------- Summary of individual treatment effects: E[Y|T=trt, X] - E[Y|T=1, X] where 'trt' is 2 and 3 2-vs-1 3-vs-1 Min. :-104.117 Min. :-207.321 1st Qu.: -37.327 1st Qu.: -61.574 Median : -4.051 Median : 6.825 Mean : -1.278 Mean : 2.121 3rd Qu.: 26.780 3rd Qu.: 68.744 Max. : 122.285 Max. : 209.722 family: gaussian loss: sq_loss_lasso method: weighting validation method: training_test_replication cutpoint: 0 replications: 2 benefit score: f_2(x): 2 vs 1, f_3(x): 3 vs 1 f_1(x): 0 minval = min(f_2(x), f_3(x)) which.min(minval) = The trt level which minimizes minval Trt recom = which.min(minval)*I(minval < c) + 1*I(minval >= c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 1 Recommended 2 Received 1 -12.4319 (SE = NA, n = 1.5) NaN (SE = NA, n = 0) Received 2 21.007 (SE = 17.8902, n = 1.5) -24.8481 (SE = 3.3736, n = 7.5) Received 3 17.1442 (SE = 0, n = 1) 35.5173 (SE = 22.6333, n = 2.5) Recommended 3 Received 1 24.4409 (SE = 4.4403, n = 7) Received 2 42.5672 (SE = NA, n = 1) Received 3 -43.2845 (SE = 7.6407, n = 3) Treatment effects conditional on subgroups: Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] -31.3354 (SE = NA, n = 4) Est of E[Y|T=2,Recom=2]-E[Y|T=/=2,Recom=2] -60.3653 (SE = 26.0069, n = 10) Est of E[Y|T=3,Recom=3]-E[Y|T=/=3,Recom=3] -72.1813 (SE = 18.3825, n = 11) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: -60.6921 (SE = 3.7457) family: gaussian loss: sq_loss_lasso method: weighting validation method: training_test_replication cutpoint: Quant_33 replications: 2 benefit score: f_2(x): 2 vs 1, f_3(x): 3 vs 1 f_1(x): 0 minval = min(f_2(x), f_3(x)) which.min(minval) = The trt level which minimizes minval Trt recom = which.min(minval)*I(minval < c) + 1*I(minval >= c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 1 Recommended 2 Received 1 0.1205 (SE = 17.7517, n = 2.5) NaN (SE = NA, n = 0) Received 2 -2.8543 (SE = 0.4226, n = 3.5) -28.3557 (SE = 3.2631, n = 5.5) Received 3 31.1302 (SE = 15.1212, n = 2) 3.6383 (SE = 8.0984, n = 1.5) Recommended 3 Received 1 27.5455 (SE = 8.8308, n = 6) Received 2 42.5672 (SE = NA, n = 1) Received 3 -43.2845 (SE = 7.6407, n = 3) Treatment effects conditional on subgroups: Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] -23.4634 (SE = 33.2635, n = 8) Est of E[Y|T=2,Recom=2]-E[Y|T=/=2,Recom=2] -31.994 (SE = 4.8353, n = 7) Est of E[Y|T=3,Recom=3]-E[Y|T=/=3,Recom=3] -73.7928 (SE = 20.6616, n = 10) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: -54.5619 (SE = 1.9297) <===============================================> family: gaussian loss: sq_loss_lasso method: weighting validation method: training_test_replication cutpoint: Quant_67 replications: 2 benefit score: f_2(x): 2 vs 1, f_3(x): 3 vs 1 f_1(x): 0 minval = min(f_2(x), f_3(x)) which.min(minval) = The trt level which minimizes minval Trt recom = which.min(minval)*I(minval < c) + 1*I(minval >= c) where c is 'cutpoint' Average Test Set Outcomes: Recommended 1 Recommended 2 Received 1 NaN (SE = NA, n = 0) NaN (SE = NA, n = 0) Received 2 NaN (SE = NA, n = 0) -18.496 (SE = 5.7219, n = 8.5) Received 3 NaN (SE = NA, n = 0) 35.5173 (SE = 22.6333, n = 2.5) Recommended 3 Received 1 19.3597 (SE = 11.6262, n = 8.5) Received 2 40.7872 (SE = NA, n = 1.5) Received 3 -6.2096 (SE = 7.7322, n = 4) Treatment effects conditional on subgroups: Est of E[Y|T=1,Recom=1]-E[Y|T=/=1,Recom=1] NaN (SE = NA, n = 0) Est of E[Y|T=2,Recom=2]-E[Y|T=/=2,Recom=2] -54.0133 (SE = 28.3552, n = 11) Est of E[Y|T=3,Recom=3]-E[Y|T=/=3,Recom=3] -29.5908 (SE = 25.0456, n = 14) Est of E[Y|Trt received = Trt recom] - E[Y|Trt received =/= Trt recom]: -42.9246 (SE = 2.8594) Hyperbolic Tangent kernel function. Hyperparameters : scale = 1 offset = 1 C 1.0000 10.0000 CV weighted accuracy 0.3839 0.3521 [ FAIL 0 | WARN 2 | SKIP 0 | PASS 238 ] [ FAIL 0 | WARN 2 | SKIP 0 | PASS 238 ] > > proc.time() user system elapsed 137.35 7.39 145.14