training/cross_validation.py

import pandas as pd
from sklearn.model_selection import StratifiedKFold
from sklearn.metrics import accuracy_score
from transformers import pipeline

# pakai model yang sama dengan sistem
model = pipeline("sentiment-analysis",
    model="w11wo/indonesian-roberta-base-sentiment-classifier")

def predict(texts):
    outputs = model(texts)
    preds = []
    for o in outputs:
        l = o['label'].lower()
        if "positive" in l:
            preds.append("positive")
        elif "negative" in l:
            preds.append("negative")
        else:
            preds.append("neutral")
    return preds


def run_cv(path="data/eval_dataset.csv", k=5):
    df = pd.read_csv(path)

    X = df["text"]
    y = df["label"]

    skf = StratifiedKFold(n_splits=k, shuffle=True, random_state=42)

    scores = []

    for train_idx, test_idx in skf.split(X, y):
        X_test = X.iloc[test_idx].tolist()
        y_test = y.iloc[test_idx].tolist()

        y_pred = predict(X_test)

        acc = accuracy_score(y_test, y_pred)
        scores.append(acc)

    print("Cross-validation scores:", scores)
    print("Mean accuracy:", sum(scores)/len(scores))


if __name__ == "__main__":
    run_cv()