Example for using exclusion constraints incorporating sums and products¶

This examples demonstrates an exclusion constraint using products and sums.

This example assumes some basic familiarity with using BayBE. We thus refer to campaign for a basic example.

Necessary imports for this example¶

import os

import numpy as np

from baybe import Campaign
from baybe.constraints import (
    DiscreteProductConstraint,
    DiscreteSumConstraint,
    ThresholdCondition,
)
from baybe.objectives import SingleTargetObjective
from baybe.parameters import (
    CategoricalParameter,
    NumericalDiscreteParameter,
    SubstanceParameter,
)
from baybe.searchspace import SearchSpace
from baybe.targets import NumericalTarget
from baybe.utils.dataframe import add_fake_results

Experiment setup¶

SMOKE_TEST = "SMOKE_TEST" in os.environ

RESOLUTION = 3 if SMOKE_TEST else 5

dict_solvent = {
    "water": "O",
    "C1": "C",
    "C2": "CC",
}
solvent = SubstanceParameter(name="Solvent", data=dict_solvent, encoding="RDKIT")
speed = CategoricalParameter(
    name="Speed", values=["slow", "normal", "fast"], encoding="INT"
)
num_parameter_1 = NumericalDiscreteParameter(
    name="NumParam1", values=list(np.linspace(0, 100, RESOLUTION)), tolerance=0.5
)
num_parameter_2 = NumericalDiscreteParameter(
    name="NumParam2", values=list(np.linspace(0, 100, RESOLUTION)), tolerance=0.5
)
num_parameter_3 = NumericalDiscreteParameter(
    name="NumParam3", values=list(np.linspace(0, 100, RESOLUTION)), tolerance=0.5
)
num_parameter_4 = NumericalDiscreteParameter(
    name="NumParam4", values=list(np.linspace(0, 100, RESOLUTION)), tolerance=0.5
)
num_parameter_5 = NumericalDiscreteParameter(
    name="NumParam5", values=list(np.linspace(0, 100, RESOLUTION)), tolerance=0.5
)
num_parameter_6 = NumericalDiscreteParameter(
    name="NumParam6", values=list(np.linspace(0, 100, RESOLUTION)), tolerance=0.5
)

parameters = [
    solvent,
    speed,
    num_parameter_1,
    num_parameter_2,
    num_parameter_3,
    num_parameter_4,
    num_parameter_5,
    num_parameter_6,
]

Creating the constraints¶

Constraints are used when creating the searchspace object. Thus, they need to be defined prior to the searchspace creation.

sum_constraint_1 = DiscreteSumConstraint(
    parameters=["NumParam1", "NumParam2"],
    condition=ThresholdCondition(threshold=150.0, operator="<="),
)
sum_constraint_2 = DiscreteSumConstraint(
    parameters=["NumParam5", "NumParam6"],
    condition=ThresholdCondition(threshold=100, operator="=", tolerance=1.0),
)
prod_constraint = DiscreteProductConstraint(
    parameters=["NumParam3", "NumParam4"],
    condition=ThresholdCondition(threshold=30, operator=">="),
)

constraints = [sum_constraint_1, sum_constraint_2, prod_constraint]

Creating the searchspace and the objective¶

searchspace = SearchSpace.from_product(parameters=parameters, constraints=constraints)

[08:37:12] DEPRECATION WARNING: please use MorganGenerator
[08:37:12] DEPRECATION WARNING: please use MorganGenerator
[08:37:12] DEPRECATION WARNING: please use MorganGenerator
[08:37:12] DEPRECATION WARNING: please use MorganGenerator
[08:37:12] DEPRECATION WARNING: please use MorganGenerator
[08:37:12] DEPRECATION WARNING: please use MorganGenerator
[08:37:12] DEPRECATION WARNING: please use MorganGenerator
[08:37:12] DEPRECATION WARNING: please use MorganGenerator
[08:37:12] DEPRECATION WARNING: please use MorganGenerator

objective = SingleTargetObjective(target=NumericalTarget(name="Target_1", mode="MAX"))

Creating and printing the campaign¶

campaign = Campaign(searchspace=searchspace, objective=objective)
print(campaign)

[1mCampaign[0m
         
 [1mMeta Data[0m
 Batches Done: 0

Fits Done: 0

 [1mSearch Space[0m
          
  [1mSearch Space Type: [0mDISCRETE
  
  [1mDiscrete Search Space[0m
               
   [1mDiscrete Parameters[0m
            Name                        Type  Num_Values                 Encoding
   0     Solvent          SubstanceParameter           3  SubstanceEncoding.RDKIT
   1       Speed        CategoricalParameter           3  CategoricalEncoding.INT
   2   NumParam1  NumericalDiscreteParameter           3                     None
   ..        ...                         ...         ...                      ...
   5   NumParam4  NumericalDiscreteParameter           3                     None
   6   NumParam5  NumericalDiscreteParameter           3                     None
   7   NumParam6  NumericalDiscreteParameter           3                     None
   
   [8 rows x 4 columns]
               
   [1mExperimental Representation[0m
               
   Solvent Speed  ...  NumParam5  NumParam6
   0     water  slow  ...        0.0      100.0
   1     water  slow  ...       50.0       50.0
   2     water  slow  ...      100.0        0.0
   ..      ...   ...  ...        ...        ...
   861      C2  fast  ...        0.0      100.0
   862      C2  fast  ...       50.0       50.0
   863      C2  fast  ...      100.0        0.0
   
   [864 rows x 8 columns]
   
   [1mMetadata:[0m

was_recommended: 0/864

was_measured: 0/864

dont_recommend: 0/864

   [1mConstraints[0m
                           Type     Affected_Parameters
   0      DiscreteSumConstraint  [NumParam1, NumParam2]
   1      DiscreteSumConstraint  [NumParam5, NumParam6]
   2  DiscreteProductConstraint  [NumParam3, NumParam4]
               
   [1mComputational Representation[0m
               
    Solvent_RDKIT_MaxAbsEStateIndex  Solvent_RDKIT_MaxPartialCharge  ...  NumParam5

NumParam6 0 0.0 -0.411510 … 0.0 100.0 1 0.0 -0.411510 … 50.0 50.0 2 0.0 -0.411510 … 100.0 0.0 .. … … … … … 861 2.0 -0.068262 … 0.0 100.0 862 2.0 -0.068262 … 50.0 50.0 863 2.0 -0.068262 … 100.0 0.0

   [864 rows x 9 columns]
 
 [1mObjective[0m
          
  [1mType: [0mSingleTargetObjective
          
  [1mTargets [0m
                Type      Name Mode  Lower_Bound  Upper_Bound Transformation
  0  NumericalTarget  Target_1  MAX         -inf          inf           None
 
 TwoPhaseMetaRecommender(initial_recommender=RandomRecommender(allow_repeated_recomm

endations=False, allow_recommending_already_measured=True), recommender=BotorchRecommender(allow_repeated_recommendations=False, allow_recommending_already_measured=True, surrogate_model=GaussianProcessSurrogate(kernel_factory=DefaultKernelFactory(), _model=None), acquisition_function=qLogExpectedImprovement(), _botorch_acqf=None, acquisition_function_cls=None, sequential_continuous=False, hybrid_sampler=None, sampling_percentage=1.0), switch_after=1)

Manual verification of the constraints¶

The following loop performs some recommendations and manually verifies the given constraints.

N_ITERATIONS = 2 if SMOKE_TEST else 5
for kIter in range(N_ITERATIONS):
    print(f"\n\n#### ITERATION {kIter+1} ####")

    print("## ASSERTS ##")
    print(
        "Number of entries with 1,2-sum above 150:      ",
        (
            campaign.searchspace.discrete.exp_rep[["NumParam1", "NumParam2"]].sum(
                axis=1
            )
            > 150.0
        ).sum(),
    )
    print(
        "Number of entries with 3,4-product under 30:   ",
        (
            campaign.searchspace.discrete.exp_rep[["NumParam3", "NumParam4"]].prod(
                axis=1
            )
            < 30
        ).sum(),
    )
    print(
        "Number of entries with 5,6-sum unequal to 100: ",
        campaign.searchspace.discrete.exp_rep[["NumParam5", "NumParam6"]]
        .sum(axis=1)
        .apply(lambda x: x - 100.0)
        .abs()
        .gt(0.01)
        .sum(),
    )

    rec = campaign.recommend(batch_size=5)
    add_fake_results(rec, campaign.targets)
    campaign.add_measurements(rec)

#### ITERATION 1 ####
## ASSERTS ##
Number of entries with 1,2-sum above 150:       0
Number of entries with 3,4-product under 30:    0
Number of entries with 5,6-sum unequal to 100:  0


#### ITERATION 2 ####
## ASSERTS ##
Number of entries with 1,2-sum above 150:       0
Number of entries with 3,4-product under 30:    0
Number of entries with 5,6-sum unequal to 100:  0