RaphaelS1/survivalmodels

[drchehidayat]

Hi Raphael

I'm opening this issue as I encountered a few problems when trying to replicate your codes in an article entitled "Neural Networks for Survival Analysis in R". Here I will share the problem for discussion here.

1. When installing Keras and TensorFlow with this code I received this error

install_keras(pip = TRUE, install_tensorflow = TRUE)
#> Error in install_keras(pip = TRUE, install_tensorflow = TRUE): could not find function "install_keras"

^{Created on 2023-04-13 with reprex v2.0.2}

ERROR: Could not find a version that satisfies the requirement tensorflow (from versions: none)
ERROR: No matching distribution found for tensorflow
Error: Error installing package(s): "'tensorflow'", "'keras'"

This error disappears when I use the codes from https://github.com/RaphaelS1/survivalmodels, install_tensorflow = FALSE instead of TRUE.

install_keras(pip = TRUE, install_tensorflow = FALSE)
#> Error in install_keras(pip = TRUE, install_tensorflow = FALSE): could not find function "install_keras"

^{Created on 2023-04-13 with reprex v2.0.2}

Then, I can proceed with the tutorial until benchmark section when I got another error when running this codes

bm <- benchmark(design)
#> Error in benchmark(design): could not find function "benchmark"

^{Created on 2023-04-13 with reprex v2.0.2}

Error: ModuleNotFoundError: No module named 'pycox'

This happened PipeOp surv.coxtime.tuned's $train()

After that, i cannot proceed with the next section Aggreggate with Harrell's C and Integrated Graf Score as there is no "bm" found.

bm$aggregate(msrs)[, c(3, 4, 7, 8)]
#> Error in eval(expr, envir, enclos): object 'bm' not found

^{Created on 2023-04-13 with reprex v2.0.2}

For your information, I'm using your package mainly "survivalmodels" and "mlr3proba" as practice for the development of survival machine learning models in my research.

I hope you help me with this. Looking forward to your answer.

Thank you.