From f47aa94d0717e4c03f11ea0ae78f63fd394d3661 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Johannes=20P=2E=20D=C3=BCrholt?=
 <johannespeter.duerholt@evonik.com>
Date: Tue, 4 Feb 2025 11:40:13 +0100
Subject: [PATCH] add randomization for fractionalfactorial

---
 .../strategies/fractional_factorial.py        |  4 ++
 bofire/strategies/fractional_factorial.py     | 16 +++++-
 tests/bofire/data_models/specs/strategies.py  |  1 +
 .../strategies/test_fractional_factorial.py   | 57 +++++++++++++++++++
 4 files changed, 75 insertions(+), 3 deletions(-)

diff --git a/bofire/data_models/strategies/fractional_factorial.py b/bofire/data_models/strategies/fractional_factorial.py
index 18b29bb45..32b9f5b4d 100644
--- a/bofire/data_models/strategies/fractional_factorial.py
+++ b/bofire/data_models/strategies/fractional_factorial.py
@@ -46,6 +46,10 @@ class FractionalFactorialStrategy(Strategy):
         int,
         Field(description="Number of reducing factors", ge=0),
     ] = 0
+    randomize_runorder: bool = Field(
+        default=False,
+        description="If true, the run order is randomized, else it is deterministic.",
+    )
 
     @classmethod
     def is_constraint_implemented(cls, my_type: Type[Constraint]) -> bool:
diff --git a/bofire/strategies/fractional_factorial.py b/bofire/strategies/fractional_factorial.py
index f3fc86f87..7c9e38342 100644
--- a/bofire/strategies/fractional_factorial.py
+++ b/bofire/strategies/fractional_factorial.py
@@ -25,6 +25,7 @@ def __init__(
         self.n_center = data_model.n_center
         self.n_generators = data_model.n_generators
         self.generator = data_model.generator
+        self.randomize_runoder = data_model.randomize_runorder
 
     def _get_continuous_design(self) -> pd.DataFrame:
         continuous_inputs = self.domain.inputs.get(ContinuousInput)
@@ -69,23 +70,32 @@ def _ask(self, candidate_count: Optional[int] = None) -> pd.DataFrame:
         if len(self.domain.inputs.get(ContinuousInput)) > 0:
             design = self._get_continuous_design()
             if len(self.domain.inputs.get(ContinuousInput)) == len(self.domain.inputs):
-                return design
+                return self.randomize_design(design)
 
         categorical_design = self._get_categorical_design()
         if len(self.domain.inputs.get([CategoricalInput, DiscreteInput])) == len(
             self.domain.inputs
         ):
-            return categorical_design
+            return self.randomize_design(categorical_design)
 
         assert isinstance(design, pd.DataFrame)
         # combine the two designs
-        return pd.concat(
+        design = pd.concat(
             [
                 pd.concat([design] * len(categorical_design), ignore_index=True),
                 pd.concat([categorical_design] * len(design), ignore_index=True),  # type: ignore
             ],
             axis=1,
         ).sort_values(by=self.domain.inputs.get_keys([CategoricalInput, DiscreteInput]))
+        return self.randomize_design(design)
+
+    def randomize_design(self, design: pd.DataFrame) -> pd.DataFrame:
+        """Randomize the run order of the design if `self.randomize_runorder` is True."""
+        return (
+            design.sample(frac=1, random_state=self._get_seed()).reset_index(drop=True)
+            if self.randomize_runoder
+            else design
+        )
 
     def has_sufficient_experiments(self) -> bool:
         return True
diff --git a/tests/bofire/data_models/specs/strategies.py b/tests/bofire/data_models/specs/strategies.py
index 0aeab2699..cadbfbe5d 100644
--- a/tests/bofire/data_models/specs/strategies.py
+++ b/tests/bofire/data_models/specs/strategies.py
@@ -589,6 +589,7 @@
         "n_center": 0,
         "n_generators": 0,
         "generator": "",
+        "randomize_runorder": False,
     },
 )
 
diff --git a/tests/bofire/strategies/test_fractional_factorial.py b/tests/bofire/strategies/test_fractional_factorial.py
index e5ce0bd0f..aa7438950 100644
--- a/tests/bofire/strategies/test_fractional_factorial.py
+++ b/tests/bofire/strategies/test_fractional_factorial.py
@@ -157,6 +157,63 @@ def test_FractionalFactorialStrategy_ask():
     assert len(candidates) == 10
 
 
+def test_FractionalFactorialStrategy_randomize_runorder():
+    # test no randomization
+    strategy_data = FractionalFactorialStrategy(
+        domain=Domain(
+            inputs=Inputs(
+                features=[
+                    ContinuousInput(key="a", bounds=(0, 1)),
+                    ContinuousInput(key="b", bounds=(-2, 8)),
+                ],
+            ),
+        ),
+        randomize_runorder=False,
+    )
+    strategy = strategies.map(strategy_data)
+    design = strategy.ask(None)
+    design2 = strategy.ask(None)
+    # test with randomization
+    assert_frame_equal(design, design2)
+    strategy_data = FractionalFactorialStrategy(
+        domain=Domain(
+            inputs=Inputs(
+                features=[
+                    ContinuousInput(key="a", bounds=(0, 1)),
+                    ContinuousInput(key="b", bounds=(-2, 8)),
+                ],
+            ),
+        ),
+        randomize_runorder=True,
+        seed=42,
+    )
+    strategy = strategies.map(strategy_data)
+    design = strategy.ask(None)
+    design2 = strategy.ask(None)
+    with pytest.raises(AssertionError):
+        assert_frame_equal(design, design2)
+    # test reproducibility with same seed for randomization
+    strategy_data = FractionalFactorialStrategy(
+        domain=Domain(
+            inputs=Inputs(
+                features=[
+                    ContinuousInput(key="a", bounds=(0, 1)),
+                    ContinuousInput(key="b", bounds=(-2, 8)),
+                ],
+            ),
+        ),
+        randomize_runorder=True,
+        seed=42,
+    )
+    strategy = strategies.map(strategy_data)
+    design3 = strategy.ask(None)
+    design4 = strategy.ask(None)
+    with pytest.raises(AssertionError):
+        assert_frame_equal(design3, design4)
+    assert_frame_equal(design, design3)
+    assert_frame_equal(design2, design4)
+
+
 def test_FractionalFactorialStrategy_ask_invalid():
     strategy_data = FractionalFactorialStrategy(
         domain=Domain(