Merge branch 'fix/clustersplit_seed' into 'dev'

Fix inconsistent results benchmarking with clustersplit

See merge request cdd/QSPRpred!177

CHANGELOG.md

@@ -6,6 +6,7 @@ From v3.0.2 to v3.0.3
 
 - Fixed a bug where an attached standardizer would be refit when calling 
 `QSPRModel.predictMols` with `use_applicability_domain=True`.
+- Fixed random seed not set in `FoldsFromDataSplit.iterFolds` for `ClusterSplit`.
 
 ## Changes
 

qsprpred/benchmarks/runner.py

@@ -158,7 +158,7 @@ def getSeedList(self, seed: int | None = None) -> list[int]:
         """
         Get a list of seeds for the replicas from one 'master' randomSeed.
         The list of seeds is generated by sampling from the range of
-        possible seeds (0 to 2^32 - 1) with the given randomSeed as the random
+        possible seeds (0 to 2**31-1) with the given randomSeed as the random
         randomSeed for the random module. This means that the list of seeds
         will be the same for each run of the benchmarking experiment
         with the same 'master' randomSeed. This is useful for reproducibility,
@@ -181,7 +181,7 @@ def getSeedList(self, seed: int | None = None) -> list[int]:
         """
         seed = seed or self.settings.random_seed
         random.seed(seed)
-        return random.sample(range(2**32 - 1), self.nRuns)
+        return random.sample(range(2**31), self.nRuns)
 
     def iterReplicas(self) -> Generator[Replica, None, None]:
         """Generator that yields `Replica` objects for each benchmarking run.

qsprpred/data/sampling/splits.py

@@ -552,3 +552,14 @@ def setSeed(self, seed: int | None):
         self.seed = seed
         if hasattr(self.clustering, "seed"):
             self.clustering.seed = seed
+
+    def getSeed(self):
+        """Get the seed for this instance.
+        
+        Returns:
+            int: the seed for this instance or None if no seed is set.
+        """
+        if hasattr(self, "seed"):
+            return self.seed
+        else:
+            return None

qsprpred/data/tables/pandas.py

@@ -112,7 +112,7 @@ def __init__(
         self.storeFormat = store_format
         self.randomState = None
         self.setRandomState(
-            random_state or int(np.random.randint(0, 2**32 - 1, dtype=np.int64))
+            random_state or int(np.random.randint(0, 2**31-1, dtype=np.int64))
         )
         self.name = name
         self.indexCols = index_cols

qsprpred/models/models.py

@@ -240,7 +240,7 @@ def initRandomState(self, random_state):
                 Random state to use for shuffling and other random operations.
         """
         if random_state is None:
-            self.randomState = int(np.random.randint(0, 2**32 - 1, dtype=np.int64))
+            self.randomState = int(np.random.randint(0, 2**31-1, dtype=np.int64))
             logger.info(
                 "No random state supplied."
                 f"Setting random state to: {self.randomState}."