inference/predict.py

import pandas as pd
#import lightgbm as lgb
#from sklearn.preprocessing import LabelEncoder
import subprocess
import sys
import warnings
warnings.filterwarnings('ignore')

try:
    import lightgbm as lgb
    from sklearn.preprocessing import LabelEncoder
    from scipy.stats import boxcox


except:
    print("An exception occurred")
    subprocess.run([sys.executable, "-m", "pip", "install", "lightgbm"])
    subprocess.run([sys.executable, "-m", "pip", "install", "scipy"])
    subprocess.run([sys.executable, "-m", "pip", "install", "scikit-learn"])
    from sklearn.preprocessing import LabelEncoder
    import lightgbm as lgb
    from scipy.stats import boxcox

    
def preprocess_input(input_data):
    # Preprocess the input data (assuming a similar preprocessing as in the training script)
    input_data['rain'] = boxcox(input_data['rain'] + 1)[0]
    return input_data

def load_model(model_path='inference/lgbm_model.txt'):
    # Load the trained LightGBM model
    model = lgb.Booster(model_file=model_path)
    return model

def predict(input_data, model):
    # Preprocess the input data
    preprocessed_data = preprocess_input(pd.DataFrame([input_data]))

    # Make predictions using the loaded model
    prediction = model.predict(preprocessed_data, num_iteration=model.best_iteration)
    predicted_class = prediction.argmax(axis=1)

    # Decode the predicted class (inverse transform)
    label_encoder = LabelEncoder()
    predicted_label = label_encoder.inverse_transform(predicted_class)[0]

    return predicted_label