scorer.py

import json
from pathlib import Path

import numpy as np
import pandas as pd
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, confusion_matrix


DATA_DIR = Path("data")
TRAIN_PATH = DATA_DIR / "train.csv"
TEST_PATH = DATA_DIR / "tester.csv"


def find_label_column(df: pd.DataFrame) -> str:
    label_cols = [c for c in df.columns if c.startswith("label_")]
    if not label_cols:
        raise ValueError("No label column found. Expected a column like label_<target_name>.")
    return sorted(label_cols)[0]


def to_int_labels(y: pd.Series) -> np.ndarray:
    if y.dtype == bool:
        return y.astype(int).to_numpy()
    if np.issubdtype(y.dtype, np.number):
        return y.astype(int).to_numpy()

    y_str = y.astype(str).str.strip().str.lower()
    mapping = {
        "0": 0, "1": 1,
        "false": 0, "true": 1,
        "no": 0, "yes": 1,
        "neg": 0, "pos": 1,
        "negative": 0, "positive": 1,
        "green": 0, "red": 1,
        "amber": 1,
    }
    if not y_str.isin(mapping.keys()).all():
        unknown = sorted(set(y_str.unique()) - set(mapping.keys()))
        raise ValueError(f"Unknown label values: {unknown}")
    return y_str.map(mapping).astype(int).to_numpy()


def main() -> None:
    if not TRAIN_PATH.exists():
        raise FileNotFoundError(f"Missing {TRAIN_PATH}")
    if not TEST_PATH.exists():
        raise FileNotFoundError(f"Missing {TEST_PATH}")

    train = pd.read_csv(TRAIN_PATH)
    test = pd.read_csv(TEST_PATH)

    label_col = find_label_column(train)
    if label_col not in test.columns:
        raise ValueError(f"Label column {label_col} missing from tester.csv")

    feature_cols = [c for c in train.columns if c != label_col]

    X_train = train[feature_cols].to_numpy(dtype=float)
    y_train = to_int_labels(train[label_col])

    X_test = test[feature_cols].to_numpy(dtype=float)
    y_test = to_int_labels(test[label_col])

    model = LogisticRegression(max_iter=2000, solver="lbfgs")
    model.fit(X_train, y_train)

    y_pred = model.predict(X_test)

    metrics = {
        "label_column": label_col,
        "n_train": int(len(train)),
        "n_test": int(len(test)),
        "accuracy": float(accuracy_score(y_test, y_pred)),
        "precision": float(precision_score(y_test, y_pred, zero_division=0)),
        "recall": float(recall_score(y_test, y_pred, zero_division=0)),
        "f1": float(f1_score(y_test, y_pred, zero_division=0)),
        "confusion_matrix": confusion_matrix(y_test, y_pred).tolist(),
    }

    print(json.dumps(metrics, indent=2))


if __name__ == "__main__":
    main()