Note
Go to the end to download the full example code.
Using ComBatGAM with MAREoS dataset#
Imports#
import warnings
import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
from sklearn.ensemble import RandomForestClassifier
from sklearn.exceptions import ConvergenceWarning
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import balanced_accuracy_score
from uniharmony import verbosity
from uniharmony.combat import ComBatGAM
from uniharmony.datasets import load_MAREoS
sns.set_theme(style="whitegrid")
verbosity("debug")
warnings.filterwarnings(action="ignore", category=ConvergenceWarning)
Data loading#
Load MAREoS benchmark dataset
datasets = load_MAREoS()loads simulated neuroimaging benchmark data.The dataset contains multiple scenarios (
truevseoseffects;simplevsinteraction; example 1/2).
# Load the MAREoS dataset (made for benchmarking harmonisation methods)
datasets = load_MAREoS()
# Define the different effects, effect types, and examples to iterate over
effects = ["true", "eos"]
effect_types = ["simple", "interaction"]
effect_examples = ["1", "2"]
random_state = 23
# Assign an empty list to each key in the results dictionary
unharmonized_results = []
neurocombat_results = []
# Define the harmonisation model to use (ComBatGAM in this case)
harm_model = ComBatGAM()
2026-05-18 13:04:40 [info ] MAREoS datasets already exist at: /home/runner/.cache/uniharmony/MAREoS
2026-05-18 13:04:40 [info ] Getting data file: /home/runner/.cache/uniharmony/MAREoS/public_datasets/eos_simple1_data.csv
2026-05-18 13:04:40 [info ] Getting data file: /home/runner/.cache/uniharmony/MAREoS/public_datasets/eos_simple2_data.csv
2026-05-18 13:04:40 [info ] Getting data file: /home/runner/.cache/uniharmony/MAREoS/public_datasets/eos_interaction1_data.csv
2026-05-18 13:04:40 [info ] Getting data file: /home/runner/.cache/uniharmony/MAREoS/public_datasets/eos_interaction2_data.csv
2026-05-18 13:04:40 [info ] Getting data file: /home/runner/.cache/uniharmony/MAREoS/public_datasets/true_simple1_data.csv
2026-05-18 13:04:40 [info ] Getting data file: /home/runner/.cache/uniharmony/MAREoS/public_datasets/true_simple2_data.csv
2026-05-18 13:04:40 [info ] Getting data file: /home/runner/.cache/uniharmony/MAREoS/public_datasets/true_interaction1_data.csv
2026-05-18 13:04:40 [info ] Getting data file: /home/runner/.cache/uniharmony/MAREoS/public_datasets/true_interaction2_data.csv
Experiments#
- Iterates all combinations:
effect=trueoreoseffect_type=simpleorinteractionexample=1or2
- For each combination:
Choose classifier: logistic regression for simple; random forest for interaction.
Extract data:
X,y,sites,folds.- Do leave-one-fold-out cross-validation:
train on folds != current fold
test on fold == current fold
Train baseline classifier on unharmonized training data and compute balanced accuracy on raw test.
Harmonize training with
NeuroComBat.fit_transform(...), then train classifier, transform test, compute balanced accuracy.
Collect results into two lists and then into DataFrames.
for effect in effects:
for e_types in effect_types:
if e_types == "interaction":
clf = RandomForestClassifier(n_estimators=10, random_state=random_state)
elif e_types == "simple":
clf = LogisticRegression(random_state=random_state)
for e_example in effect_examples:
example = effect + "_" + e_types + e_example
print(f"Running experiment: {example}")
data = datasets[example]
sites = data["sites"]
X = data["X"]
folds = data["folds"]
folds = pd.Series(folds)
sites = data["sites"]
target = data["y"]
covars = target.ravel().reshape(-1, 1)
for fold in folds.unique():
# Train Data
X = data["X"].copy()
y = data["y"].copy()
sites = data["sites"].copy()
# Train Target
X_train = X[data["folds"] != fold]
site_train = sites[data["folds"] != fold]
y_train = y[data["folds"] != fold]
# Test data
X_test = X[data["folds"] == fold]
site_test = sites[data["folds"] == fold]
# Test target
y_test = y[data["folds"] == fold]
# Unharmonized baseline model
clf.fit(X_train, y_train)
unharmonized_results.append(
[
balanced_accuracy_score(y_true=y_test, y_pred=clf.predict(X=X_test)),
fold,
effect,
e_types,
e_example,
example,
]
)
# ComBatGAM (do not include target as covariate - avoiding data leakage)
X_train_harm = harm_model.fit_transform(X=X_train, sites=site_train, smooth_covariates=y_train.reshape(-1, 1))
# Fit the model with the harmonized train
clf.fit(X_train_harm, y_train)
# harmonize the test data
X_test_harm = harm_model.transform(X=X_test, sites=site_test, smooth_covariates=y_test.reshape(-1, 1))
neurocombat_results.append(
[
balanced_accuracy_score(y_true=y_test, y_pred=clf.predict(X=X_test_harm)),
fold,
effect,
e_types,
e_example,
example,
]
)
# Results to dataframe
unharmonized_results = pd.DataFrame(data=unharmonized_results, columns=["bACC", "Fold", "Effect", "Type", "Example", "Name"])
unharmonized_results["Method"] = "Unharmonized Baseline"
neurocombat_results = pd.DataFrame(data=neurocombat_results, columns=["bACC", "Fold", "Effect", "Type", "Example", "Name"])
neurocombat_results["Method"] = "ComBatGAM"
results = pd.concat([unharmonized_results, neurocombat_results])
Running experiment: true_simple1
2026-05-18 13:04:40 [debug ] Fitting
2026-05-18 13:04:40 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:04:40 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:04:40 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:05:04 [debug ] Transforming
2026-05-18 13:05:04 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:05:04 [debug ] Transforming
2026-05-18 13:05:04 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:05:04 [debug ] Fitting
2026-05-18 13:05:04 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:05:04 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:05:04 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:05:26 [debug ] Transforming
2026-05-18 13:05:26 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:05:26 [debug ] Transforming
2026-05-18 13:05:26 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:05:26 [debug ] Fitting
2026-05-18 13:05:26 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:05:26 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:05:26 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:05:48 [debug ] Transforming
2026-05-18 13:05:48 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:05:48 [debug ] Transforming
2026-05-18 13:05:48 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:05:48 [debug ] Fitting
2026-05-18 13:05:48 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:05:48 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:05:48 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:06:10 [debug ] Transforming
2026-05-18 13:06:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:06:10 [debug ] Transforming
2026-05-18 13:06:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:06:10 [debug ] Fitting
2026-05-18 13:06:10 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:06:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:06:10 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:06:32 [debug ] Transforming
2026-05-18 13:06:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:06:32 [debug ] Transforming
2026-05-18 13:06:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:06:32 [debug ] Fitting
2026-05-18 13:06:32 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:06:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:06:32 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:06:54 [debug ] Transforming
2026-05-18 13:06:54 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:06:54 [debug ] Transforming
2026-05-18 13:06:54 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:06:54 [debug ] Fitting
2026-05-18 13:06:54 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:06:54 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:06:54 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:07:16 [debug ] Transforming
2026-05-18 13:07:16 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:07:16 [debug ] Transforming
2026-05-18 13:07:16 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:07:16 [debug ] Fitting
2026-05-18 13:07:16 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:07:16 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:07:16 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:07:38 [debug ] Transforming
2026-05-18 13:07:38 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:07:38 [debug ] Transforming
2026-05-18 13:07:38 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:07:38 [debug ] Fitting
2026-05-18 13:07:38 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:07:38 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:07:38 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:08:00 [debug ] Transforming
2026-05-18 13:08:00 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:08:00 [debug ] Transforming
2026-05-18 13:08:00 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:08:00 [debug ] Fitting
2026-05-18 13:08:00 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:08:00 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:08:00 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:08:22 [debug ] Transforming
2026-05-18 13:08:22 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:08:22 [debug ] Transforming
2026-05-18 13:08:22 [debug ] Setting up smoothing using B-Splines
Running experiment: true_simple2
2026-05-18 13:08:22 [debug ] Fitting
2026-05-18 13:08:22 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:08:22 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:08:22 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:08:44 [debug ] Transforming
2026-05-18 13:08:44 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:08:44 [debug ] Transforming
2026-05-18 13:08:44 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:08:44 [debug ] Fitting
2026-05-18 13:08:44 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:08:44 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:08:44 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:09:07 [debug ] Transforming
2026-05-18 13:09:07 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:09:07 [debug ] Transforming
2026-05-18 13:09:07 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:09:07 [debug ] Fitting
2026-05-18 13:09:07 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:09:07 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:09:07 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:09:29 [debug ] Transforming
2026-05-18 13:09:29 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:09:29 [debug ] Transforming
2026-05-18 13:09:29 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:09:29 [debug ] Fitting
2026-05-18 13:09:29 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:09:29 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:09:29 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:09:52 [debug ] Transforming
2026-05-18 13:09:52 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:09:52 [debug ] Transforming
2026-05-18 13:09:52 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:09:52 [debug ] Fitting
2026-05-18 13:09:52 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:09:52 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:09:52 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:10:14 [debug ] Transforming
2026-05-18 13:10:14 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:10:14 [debug ] Transforming
2026-05-18 13:10:14 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:10:14 [debug ] Fitting
2026-05-18 13:10:14 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:10:14 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:10:14 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:10:36 [debug ] Transforming
2026-05-18 13:10:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:10:36 [debug ] Transforming
2026-05-18 13:10:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:10:36 [debug ] Fitting
2026-05-18 13:10:36 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:10:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:10:36 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:10:58 [debug ] Transforming
2026-05-18 13:10:58 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:10:58 [debug ] Transforming
2026-05-18 13:10:58 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:10:58 [debug ] Fitting
2026-05-18 13:10:58 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:10:58 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:10:58 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:11:19 [debug ] Transforming
2026-05-18 13:11:19 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:11:19 [debug ] Transforming
2026-05-18 13:11:19 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:11:19 [debug ] Fitting
2026-05-18 13:11:19 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:11:19 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:11:19 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:11:41 [debug ] Transforming
2026-05-18 13:11:41 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:11:41 [debug ] Transforming
2026-05-18 13:11:41 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:11:41 [debug ] Fitting
2026-05-18 13:11:41 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:11:41 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:11:41 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:12:03 [debug ] Transforming
2026-05-18 13:12:03 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:12:03 [debug ] Transforming
2026-05-18 13:12:03 [debug ] Setting up smoothing using B-Splines
Running experiment: true_interaction1
2026-05-18 13:12:03 [debug ] Fitting
2026-05-18 13:12:03 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:12:03 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:12:03 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:12:24 [debug ] Transforming
2026-05-18 13:12:24 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:12:24 [debug ] Transforming
2026-05-18 13:12:24 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:12:24 [debug ] Fitting
2026-05-18 13:12:24 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:12:24 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:12:24 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:12:45 [debug ] Transforming
2026-05-18 13:12:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:12:45 [debug ] Transforming
2026-05-18 13:12:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:12:45 [debug ] Fitting
2026-05-18 13:12:45 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:12:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:12:45 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:13:06 [debug ] Transforming
2026-05-18 13:13:06 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:13:06 [debug ] Transforming
2026-05-18 13:13:06 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:13:06 [debug ] Fitting
2026-05-18 13:13:06 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:13:06 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:13:06 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:13:28 [debug ] Transforming
2026-05-18 13:13:28 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:13:28 [debug ] Transforming
2026-05-18 13:13:28 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:13:28 [debug ] Fitting
2026-05-18 13:13:28 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:13:28 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:13:28 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:13:49 [debug ] Transforming
2026-05-18 13:13:49 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:13:49 [debug ] Transforming
2026-05-18 13:13:49 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:13:49 [debug ] Fitting
2026-05-18 13:13:49 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:13:49 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:13:49 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:14:10 [debug ] Transforming
2026-05-18 13:14:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:14:10 [debug ] Transforming
2026-05-18 13:14:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:14:10 [debug ] Fitting
2026-05-18 13:14:10 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:14:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:14:10 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:14:32 [debug ] Transforming
2026-05-18 13:14:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:14:32 [debug ] Transforming
2026-05-18 13:14:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:14:32 [debug ] Fitting
2026-05-18 13:14:32 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:14:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:14:32 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:14:53 [debug ] Transforming
2026-05-18 13:14:53 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:14:53 [debug ] Transforming
2026-05-18 13:14:53 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:14:53 [debug ] Fitting
2026-05-18 13:14:53 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:14:53 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:14:53 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:15:14 [debug ] Transforming
2026-05-18 13:15:14 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:15:14 [debug ] Transforming
2026-05-18 13:15:14 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:15:14 [debug ] Fitting
2026-05-18 13:15:14 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:15:14 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:15:14 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:15:36 [debug ] Transforming
2026-05-18 13:15:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:15:36 [debug ] Transforming
2026-05-18 13:15:36 [debug ] Setting up smoothing using B-Splines
Running experiment: true_interaction2
2026-05-18 13:15:36 [debug ] Fitting
2026-05-18 13:15:36 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:15:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:15:36 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:15:58 [debug ] Transforming
2026-05-18 13:15:58 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:15:58 [debug ] Transforming
2026-05-18 13:15:58 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:15:58 [debug ] Fitting
2026-05-18 13:15:58 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:15:58 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:15:58 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:16:19 [debug ] Transforming
2026-05-18 13:16:19 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:16:19 [debug ] Transforming
2026-05-18 13:16:19 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:16:19 [debug ] Fitting
2026-05-18 13:16:19 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:16:19 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:16:19 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:16:41 [debug ] Transforming
2026-05-18 13:16:41 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:16:41 [debug ] Transforming
2026-05-18 13:16:41 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:16:41 [debug ] Fitting
2026-05-18 13:16:41 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:16:41 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:16:41 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:17:03 [debug ] Transforming
2026-05-18 13:17:03 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:17:03 [debug ] Transforming
2026-05-18 13:17:03 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:17:03 [debug ] Fitting
2026-05-18 13:17:03 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:17:03 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:17:03 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:17:25 [debug ] Transforming
2026-05-18 13:17:25 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:17:25 [debug ] Transforming
2026-05-18 13:17:25 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:17:25 [debug ] Fitting
2026-05-18 13:17:25 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:17:25 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:17:25 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:17:46 [debug ] Transforming
2026-05-18 13:17:46 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:17:46 [debug ] Transforming
2026-05-18 13:17:46 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:17:46 [debug ] Fitting
2026-05-18 13:17:46 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:17:46 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:17:46 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:18:08 [debug ] Transforming
2026-05-18 13:18:08 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:18:08 [debug ] Transforming
2026-05-18 13:18:08 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:18:08 [debug ] Fitting
2026-05-18 13:18:08 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:18:08 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:18:08 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:18:29 [debug ] Transforming
2026-05-18 13:18:29 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:18:29 [debug ] Transforming
2026-05-18 13:18:29 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:18:29 [debug ] Fitting
2026-05-18 13:18:29 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:18:29 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:18:29 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:18:51 [debug ] Transforming
2026-05-18 13:18:51 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:18:51 [debug ] Transforming
2026-05-18 13:18:51 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:18:51 [debug ] Fitting
2026-05-18 13:18:51 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:18:51 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:18:51 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:13 [debug ] Transforming
2026-05-18 13:19:13 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:13 [debug ] Transforming
2026-05-18 13:19:13 [debug ] Setting up smoothing using B-Splines
Running experiment: eos_simple1
2026-05-18 13:19:13 [debug ] Fitting
2026-05-18 13:19:13 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:13 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:13 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:18 [debug ] Transforming
2026-05-18 13:19:18 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:18 [debug ] Transforming
2026-05-18 13:19:18 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:18 [debug ] Fitting
2026-05-18 13:19:18 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:18 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:18 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:22 [debug ] Transforming
2026-05-18 13:19:22 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:22 [debug ] Transforming
2026-05-18 13:19:22 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:22 [debug ] Fitting
2026-05-18 13:19:22 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:22 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:22 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:27 [debug ] Transforming
2026-05-18 13:19:27 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:27 [debug ] Transforming
2026-05-18 13:19:27 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:27 [debug ] Fitting
2026-05-18 13:19:27 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:27 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:27 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:31 [debug ] Transforming
2026-05-18 13:19:31 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:31 [debug ] Transforming
2026-05-18 13:19:31 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:31 [debug ] Fitting
2026-05-18 13:19:31 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:31 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:31 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:36 [debug ] Transforming
2026-05-18 13:19:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:36 [debug ] Transforming
2026-05-18 13:19:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:36 [debug ] Fitting
2026-05-18 13:19:36 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:36 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:40 [debug ] Transforming
2026-05-18 13:19:40 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:40 [debug ] Transforming
2026-05-18 13:19:40 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:40 [debug ] Fitting
2026-05-18 13:19:40 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:40 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:40 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:45 [debug ] Transforming
2026-05-18 13:19:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:45 [debug ] Transforming
2026-05-18 13:19:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:45 [debug ] Fitting
2026-05-18 13:19:45 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:45 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:49 [debug ] Transforming
2026-05-18 13:19:49 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:49 [debug ] Transforming
2026-05-18 13:19:49 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:49 [debug ] Fitting
2026-05-18 13:19:49 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:49 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:49 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:54 [debug ] Transforming
2026-05-18 13:19:54 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:54 [debug ] Transforming
2026-05-18 13:19:54 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:54 [debug ] Fitting
2026-05-18 13:19:54 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:54 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:54 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:19:58 [debug ] Transforming
2026-05-18 13:19:58 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:58 [debug ] Transforming
2026-05-18 13:19:58 [debug ] Setting up smoothing using B-Splines
Running experiment: eos_simple2
2026-05-18 13:19:58 [debug ] Fitting
2026-05-18 13:19:58 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:19:58 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:19:58 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:20:20 [debug ] Transforming
2026-05-18 13:20:20 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:20:20 [debug ] Transforming
2026-05-18 13:20:20 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:20:20 [debug ] Fitting
2026-05-18 13:20:20 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:20:20 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:20:20 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:20:42 [debug ] Transforming
2026-05-18 13:20:42 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:20:42 [debug ] Transforming
2026-05-18 13:20:42 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:20:42 [debug ] Fitting
2026-05-18 13:20:42 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:20:42 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:20:42 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:21:04 [debug ] Transforming
2026-05-18 13:21:04 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:21:04 [debug ] Transforming
2026-05-18 13:21:04 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:21:04 [debug ] Fitting
2026-05-18 13:21:04 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:21:04 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:21:04 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:21:26 [debug ] Transforming
2026-05-18 13:21:26 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:21:26 [debug ] Transforming
2026-05-18 13:21:26 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:21:26 [debug ] Fitting
2026-05-18 13:21:26 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:21:26 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:21:26 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:21:48 [debug ] Transforming
2026-05-18 13:21:48 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:21:48 [debug ] Transforming
2026-05-18 13:21:48 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:21:48 [debug ] Fitting
2026-05-18 13:21:48 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:21:48 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:21:48 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:22:10 [debug ] Transforming
2026-05-18 13:22:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:22:10 [debug ] Transforming
2026-05-18 13:22:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:22:10 [debug ] Fitting
2026-05-18 13:22:10 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:22:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:22:10 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:22:32 [debug ] Transforming
2026-05-18 13:22:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:22:32 [debug ] Transforming
2026-05-18 13:22:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:22:32 [debug ] Fitting
2026-05-18 13:22:32 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:22:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:22:32 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:22:54 [debug ] Transforming
2026-05-18 13:22:54 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:22:54 [debug ] Transforming
2026-05-18 13:22:54 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:22:54 [debug ] Fitting
2026-05-18 13:22:54 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:22:54 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:22:54 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:23:16 [debug ] Transforming
2026-05-18 13:23:16 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:23:16 [debug ] Transforming
2026-05-18 13:23:16 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:23:16 [debug ] Fitting
2026-05-18 13:23:16 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:23:16 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:23:16 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:23:38 [debug ] Transforming
2026-05-18 13:23:38 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:23:38 [debug ] Transforming
2026-05-18 13:23:38 [debug ] Setting up smoothing using B-Splines
Running experiment: eos_interaction1
2026-05-18 13:23:38 [debug ] Fitting
2026-05-18 13:23:38 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:23:38 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:23:38 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:23:59 [debug ] Transforming
2026-05-18 13:23:59 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:23:59 [debug ] Transforming
2026-05-18 13:23:59 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:24:00 [debug ] Fitting
2026-05-18 13:24:00 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:24:00 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:24:00 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:24:21 [debug ] Transforming
2026-05-18 13:24:21 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:24:21 [debug ] Transforming
2026-05-18 13:24:21 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:24:21 [debug ] Fitting
2026-05-18 13:24:21 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:24:21 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:24:21 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:24:42 [debug ] Transforming
2026-05-18 13:24:42 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:24:42 [debug ] Transforming
2026-05-18 13:24:42 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:24:42 [debug ] Fitting
2026-05-18 13:24:42 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:24:42 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:24:42 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:25:03 [debug ] Transforming
2026-05-18 13:25:03 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:25:03 [debug ] Transforming
2026-05-18 13:25:03 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:25:03 [debug ] Fitting
2026-05-18 13:25:03 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:25:03 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:25:03 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:25:24 [debug ] Transforming
2026-05-18 13:25:24 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:25:24 [debug ] Transforming
2026-05-18 13:25:24 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:25:24 [debug ] Fitting
2026-05-18 13:25:24 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:25:24 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:25:24 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:25:45 [debug ] Transforming
2026-05-18 13:25:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:25:45 [debug ] Transforming
2026-05-18 13:25:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:25:45 [debug ] Fitting
2026-05-18 13:25:45 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:25:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:25:45 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:26:06 [debug ] Transforming
2026-05-18 13:26:06 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:26:06 [debug ] Transforming
2026-05-18 13:26:06 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:26:06 [debug ] Fitting
2026-05-18 13:26:06 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:26:06 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:26:06 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:26:27 [debug ] Transforming
2026-05-18 13:26:27 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:26:28 [debug ] Transforming
2026-05-18 13:26:28 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:26:28 [debug ] Fitting
2026-05-18 13:26:28 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:26:28 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:26:28 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:26:49 [debug ] Transforming
2026-05-18 13:26:49 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:26:49 [debug ] Transforming
2026-05-18 13:26:49 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:26:49 [debug ] Fitting
2026-05-18 13:26:49 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:26:49 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:26:49 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:10 [debug ] Transforming
2026-05-18 13:27:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:10 [debug ] Transforming
2026-05-18 13:27:10 [debug ] Setting up smoothing using B-Splines
Running experiment: eos_interaction2
2026-05-18 13:27:10 [debug ] Fitting
2026-05-18 13:27:10 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:27:10 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:10 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:14 [debug ] Transforming
2026-05-18 13:27:14 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:14 [debug ] Transforming
2026-05-18 13:27:14 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:14 [debug ] Fitting
2026-05-18 13:27:14 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:27:14 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:14 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:19 [debug ] Transforming
2026-05-18 13:27:19 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:19 [debug ] Transforming
2026-05-18 13:27:19 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:19 [debug ] Fitting
2026-05-18 13:27:19 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:27:19 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:19 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:23 [debug ] Transforming
2026-05-18 13:27:23 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:23 [debug ] Transforming
2026-05-18 13:27:23 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:23 [debug ] Fitting
2026-05-18 13:27:23 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:27:23 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:23 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:27 [debug ] Transforming
2026-05-18 13:27:27 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:27 [debug ] Transforming
2026-05-18 13:27:27 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:28 [debug ] Fitting
2026-05-18 13:27:28 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:27:28 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:28 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:32 [debug ] Transforming
2026-05-18 13:27:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:32 [debug ] Transforming
2026-05-18 13:27:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:32 [debug ] Fitting
2026-05-18 13:27:32 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:27:32 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:32 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:36 [debug ] Transforming
2026-05-18 13:27:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:36 [debug ] Transforming
2026-05-18 13:27:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:36 [debug ] Fitting
2026-05-18 13:27:36 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:27:36 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:36 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:41 [debug ] Transforming
2026-05-18 13:27:41 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:41 [debug ] Transforming
2026-05-18 13:27:41 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:41 [debug ] Fitting
2026-05-18 13:27:41 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:27:41 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:41 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:45 [debug ] Transforming
2026-05-18 13:27:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:45 [debug ] Transforming
2026-05-18 13:27:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:45 [debug ] Fitting
2026-05-18 13:27:45 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:27:45 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:45 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:49 [debug ] Transforming
2026-05-18 13:27:49 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:50 [debug ] Transforming
2026-05-18 13:27:50 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:50 [debug ] Fitting
2026-05-18 13:27:50 [info ] If you intend to build a machine learning (ML) model,then make sure that you DO *NOT* preserve the ML model's target as covariate. You will be required to provide the covariate also at transform time, and this will produce data leakage. If you are performing a statistical analysis and want to preserve a variable of interest, then it is correct to specify it as covariate.
2026-05-18 13:27:50 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:50 [debug ] Final formula for smoothing: y ~ x0 + x1 + x2 + x3 + x4 + x5 + x6 + x7 - 1
2026-05-18 13:27:54 [debug ] Transforming
2026-05-18 13:27:54 [debug ] Setting up smoothing using B-Splines
2026-05-18 13:27:54 [debug ] Transforming
2026-05-18 13:27:54 [debug ] Setting up smoothing using B-Splines
Plotting#
fig, ax = plt.subplots(1, 1, figsize=[15, 7])
harm_methods = [
"ComBatGAM",
"Unharmonized Baseline",
]
sns.swarmplot(data=results, x="Name", y="bACC", hue="Method", hue_order=harm_methods, dodge=True, ax=ax)
sns.boxplot(
data=results,
color="w",
zorder=1,
x="Name",
y="bACC",
hue="Method",
hue_order=harm_methods,
dodge=True,
ax=ax,
palette=["w"] * len(harm_methods),
)
handles, labels = ax.get_legend_handles_labels()
ax.legend(handles[: len(harm_methods)], labels[: len(harm_methods)])
ax.axhline(0.5, lw=2, color="k", ls="--", alpha=0.7, label="Chance level")
plt.grid(axis="y")
plt.grid(axis="y")
plt.xticks(rotation=45)
plt.show()
Take-home message
ComBatGAM removes the eos site effect while preserving the “true” signal.
The plot and results demonstrate the method’s ability to reduce spurious
site-related variation and maintain true biological effect performance.
Total running time of the script: (23 minutes 14.802 seconds)