From 5edf6ecb55acf3cb916a8fb17861fd5c4bf1d58a Mon Sep 17 00:00:00 2001
From: lewismervin1 <58475084+lewismervin1@users.noreply.github.com>
Date: Wed, 3 Jul 2024 08:55:47 +0100
Subject: [PATCH] Update optbuild.py

remove AZ specific inference function
---
 optunaz/optbuild.py | 50 ---------------------------------------------
 1 file changed, 50 deletions(-)

diff --git a/optunaz/optbuild.py b/optunaz/optbuild.py
index 3ef0ad7..eec9634 100644
--- a/optunaz/optbuild.py
+++ b/optunaz/optbuild.py
@@ -22,44 +22,6 @@
 logger = logging.getLogger(__name__)
 
 
-def predict_pls(model_path, inference_path):
-    if inference_path == "None":
-        logger.info(f"Inference path is not set so AL predictions not performed")
-        return
-    else:
-        logger.info(f"Inference path is {inference_path}")
-    predict_args = [
-        "prog",
-        "--model-file",
-        str(model_path),
-        "--input-smiles-csv-file",
-        str(inference_path),
-        "--input-smiles-csv-column",
-        "Structure",
-        "--output-prediction-csv-file",
-        str(os.path.dirname(model_path)) + "/al.csv",
-        "--predict-uncertainty",
-        "--uncertainty_quantile",
-        "0.99",
-    ]
-    try:
-        with patch.object(sys, "argv", predict_args):
-            logging.info("Performing active learning predictions")
-            predict.main()
-    except FileNotFoundError:
-        logger.info(
-            f"PLS file not found at {model_path}, AL predictions not performed"
-        )
-    except predict.UncertaintyError:
-        logging.info(
-            "PLS prediction not performed: algorithm does not support uncertainty prediction"
-        )
-    except predict.AuxCovariateMissing:
-        logging.info(
-            "PLS prediction not performed: algorithm requires corvariate auxiliary data for inference"
-        )
-
-
 def main():
     logging.basicConfig(level=logging.INFO)
     parser = argparse.ArgumentParser(
@@ -97,12 +59,6 @@ def main():
         help="Turn off descriptor generation caching ",
         action="store_true",
     )
-    parser.add_argument(
-        "--inference_uncert",
-        help="Path for uncertainty inference and thresholding.",
-        type=pathlib.Path,
-        default="/projects/db-mirror/MLDatasets/PLS/pls.csv",
-    )
     args = parser.parse_args()
 
     AnyConfig = Union[OptimizationConfig, BuildConfig]
@@ -111,7 +67,6 @@ def main():
 
     if isinstance(config, OptimizationConfig):
         study_name = str(pathlib.Path(args.config).absolute())
-        pred_pls = False
         if not args.no_cache:
             config.set_cache()
             cache = config._cache
@@ -123,7 +78,6 @@ def main():
         if args.best_model_outpath or args.merged_model_outpath:
             buildconfig = buildconfig_best(study)
     elif isinstance(config, BuildConfig):
-        pred_pls = True
         buildconfig = config
         cache = None
         cache_dir = None
@@ -140,15 +94,11 @@ def main():
             args.best_model_outpath,
             cache=cache,
         )
-        if not args.merged_model_outpath and pred_pls:
-            predict_pls(args.best_model_outpath, args.inference_uncert)
     if args.merged_model_outpath:
         build_merged(
             buildconfig,
             args.merged_model_outpath,
             cache=cache,
         )
-        if pred_pls:
-            predict_pls(args.merged_model_outpath, args.inference_uncert)
     if cache_dir is not None:
         cache_dir.cleanup()