..

README.md

@@ -12,7 +12,12 @@ pip install HEBO==0.3.5 --no-deps
 Experiments:
 
 * create experiments scripts and run them (`run_experiments.sh`), e.g., via slurm submit scripts
-* this will create folders in `results/`
+* this will create folders and result files in `results/`
+
+Below is example code to generate experiments (e.g., for CatBoost).
+See `main.py` and the main logic in `reshufflebench`.
+Code in `analyze/` is used to analyze experiment results.
+Code in `visualize` is used to visualize analyzed experiment results.
 
 Random Holdout
 

simulations/README.md

@@ -11,4 +11,4 @@ Simulations:
 `python create_simulations.py`
 
 Afterwards, analyses of these result files can be performed via `analyze.R` to generate figures in `plots/`.
-
+You can obtain raw results via the following link: `https://www.dropbox.com/scl/fi/2ql036q33ftdjoff2jv8u/simulation_results.zip?rlkey=y81yv6ynejubmrls1jfwt4qns&st=rjnh6esb&dl=0` and unzip these csvs and place them in `results/`.

simulations/simulate.py

@@ -107,7 +107,6 @@ def kernel_factory(
     torch.manual_seed(42)
 
     results_path = "results"
-    plots_path = "plots"
     d = 1
     alpha = args.alpha
     lengthscale = args.lengthscale