sparse_gp_wtdata_czc_nkmeans.py

#!/usr/bin/env python

# Standard libraries
import time
import numpy as np
import warnings
import csv

from sparse_gp_wtdata import *

warnings.filterwarnings("ignore")


def save_czc_nkmeans_results(results):
    if results:
        fields = results[0].keys()
        with open(f"sgp_wtdata_results_czc_nkmeans.csv", "w", newline="") as file:
            writer = csv.DictWriter(file, delimiter=",", fieldnames=fields)
            writer.writeheader()
            writer.writerows(results)


if __name__ == "__main__":
    start = time.time()
    print("Loading data...")
    df = load_database()
    X = np.array(df[INPUT_NAMES])
    print("Data loaded in {:.2f}s".format(time.time() - start))

    # Set seed
    np.random.seed(0)

    print("Computing...")
    start = time.time()

    results = []

    inducing_method = "NORM_KMEANS"
    output_name = "CZC"
    Y = np.array(df[[output_name]])

    for M in [50, 100, 200, 500, 1000]:
        for sparse_method in SPARSE_METHODS:
            print(
                f"*** {output_name} - {sparse_method} - {inducing_method} - M={M} ******************"
            )
            res = sgp_compute(X, Y, output_name, sparse_method, inducing_method, M)
            print(res)
            results.append(res)

            # save intermediate results
            save_czc_nkmeans_results(results)

    elapsed = time.time() - start
    print("Computation in {:.2f}s".format(time.time() - start))
    save_czc_nkmeans_results(results)