Getting Started with trustcv

download Installation

The base package includes scikit-learn integration and all core cross-validation methods. Install optional extras for deep learning framework support.

Optional Extras

git clone https://github.com/ki-smile/trustcv.git
cd trustcv
pip install -e .

rocket_launch Quick Start — Basic Example

Load a dataset and create a scikit-learn pipeline, then validate it with a single call using TrustCVValidator.

1 Prepare your model

from sklearn.datasets import load_breast_cancer
from sklearn.pipeline import make_pipeline
from sklearn.preprocessing import StandardScaler
from sklearn.ensemble import RandomForestClassifier

X, y = load_breast_cancer(return_X_y=True)
model = make_pipeline(StandardScaler(), RandomForestClassifier(random_state=42))

2 Validate with TrustCVValidator

from trustcv import TrustCVValidator

validator = TrustCVValidator(
    method="stratified_kfold",
    n_splits=5,
    metrics=["accuracy", "roc_auc", "f1"],
    check_leakage=True,
    check_balance=True,
    return_confidence_intervals=True,
)

val_result = validator.validate(model=model, X=X, y=y)
print(val_result.summary())

settings_suggest Using UniversalCVRunner

For more control over the cross-validation strategy, use UniversalCVRunner with any splitter from the library. It automatically detects your framework.

from trustcv import StratifiedKFold, UniversalCVRunner

cv = StratifiedKFold(n_splits=5, shuffle=True, random_state=42)
runner = UniversalCVRunner(cv_splitter=cv, framework="auto")

results = runner.run(model=model, data=(X, y), metrics=["accuracy", "roc_auc"])
print(results.summary())

verified_user Data Integrity Checks

Before training, detect potential data leakage and check class balance to ensure your evaluation is trustworthy.

from trustcv import DataLeakageChecker, BalanceChecker

leak_report = DataLeakageChecker().check(X=X, y=y, n_splits=5, random_state=42)
balance_report = BalanceChecker().check_class_balance(y)

print(leak_report.summary)
print(balance_report)

explore Next Steps

Now that you have trustcv installed, explore these resources to get the most out of the toolkit.