fincausal 0.2.0

Finance decision intelligence tools for cost-sensitive credit and fraud risk modelling.

FinCausal

FinCausal is a finance-focused Python library for decision intelligence in credit and fraud risk workflows.

It helps analysts move from:

"This model has good accuracy"

to:

"This decision threshold reduces expected financial loss under these cost assumptions."

Version 0.2.0 focus

• explicit CostMatrix object
• threshold optimization by expected financial loss
• professional finance-style decision reports
• stress-test summaries
• model-based intervention simulation
• stronger input validation
• scikit-learn-style estimator design
• synthetic credit-risk dataset for demos and tests

Install

pip install fincausal

Quickstart

from sklearn.model_selection import train_test_split
from fincausal import CostMatrix, CostSensitiveClassifier, make_credit_risk_data

X, y = make_credit_risk_data(n_samples=3000, random_state=42)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42, stratify=y)

costs = CostMatrix(false_positive=50, false_negative=500, currency="GBP")
model = CostSensitiveClassifier(cost_matrix=costs)
model.fit(X_train, y_train)
model.optimize_threshold(X_test, y_test)
print(model.summary(X_test, y_test))

Current status

This is an early open-source project. It is not a production-grade bank model-risk platform.

Roadmap

• SHAP-based explanation wrapper
• HTML/PDF report export
• model-card templates
• fraud-risk dataset demo
• MLflow experiment logging
• validation and monitoring reports
• documentation website

License

MIT License.