import numpy as np
from models import MLP, SEED, train_test_split
dataset = np.genfromtxt('data_points.csv', delimiter=',')
x = dataset[1:, :2]
y = dataset[1:, 2]
y = np.expand_dims(y, -1) # row -> column vector
# train 70%, test 30%
x_train, x_test, y_train, y_test = train_test_split(x,
y,
train_size=0.7,
random_state=SEED)
mlp = MLP(hidden_size=25, _rho=1e-5)
mlp.extreme_learning(x_train, y_train)
print(f'Test error: {mlp.test_loss(x_test, y_test):.4f}')
mlp.surface_plot(x_test)
