def test_handle_errors(self):
with self.assertRaises(ValueError):
# Wrong: size of target data not the same as size of
# input data.
GRNN().train(
np.array([[0], [0]]), np.array([0])
)
with self.assertRaises(ValueError):
# Wrong: 2-D target vector (must be 1-D)
GRNN().train(
np.array([[0], [0]]), np.array([[0]])
)
with self.assertRaises(AttributeError):
# Wrong: can't use iterative learning process for this
# algorithm
GRNN().train_epoch()
with self.assertRaises(ValueError):
# Wrong: invalid feature size for prediction data
grnet = GRNN()
grnet.train(np.array([[0], [0]]), np.array([0]))
grnet.predict(np.array([[0]]))
def build_model2(std, train_size, t, x):
train_size = int(t.shape[0] * train_size)
# X_train = t[:train_size]
# y_train = x[:train_size]
# X_test = t[train_size:]
# y_test = x[train_size:]
X_train, X_test, y_train, y_test = train_test_split(t,
x,
train_size=train_size,
shuffle=True,
random_state=14)
scaler_x = StandardScaler()
scaler_y = StandardScaler()
tmp_train_scaled_x = scaler_x.fit_transform(X_train[:, np.newaxis])
tmp_test_scaled_x = scaler_x.transform(X_test[:, np.newaxis])
tmp_train_scaled_y = scaler_y.fit_transform(y_train[:, np.newaxis])
grnn = GRNN(std=std)
grnn.fit(tmp_train_scaled_x, tmp_train_scaled_y)
pred_x = grnn.predict(tmp_train_scaled_x)
pred_x = scaler_y.inverse_transform(pred_x)
mse = mean_squared_error(y_train, pred_x.flatten())
print(f'RMSE = {np.sqrt(mse)}')
plt.plot(t, x, c='r')
plt.scatter(X_train, y_train, label='train')
plt.scatter(X_train, pred_x, label='predict')
plt.legend()
plt.show()
pred_x = grnn.predict(tmp_test_scaled_x)
pred_x = scaler_y.inverse_transform(pred_x)
mse = mean_squared_error(y_test, pred_x.flatten())
print(f'RMSE = {np.sqrt(mse)}')
plt.plot(t, x, c='r')
plt.scatter(X_test, y_test, label='test')
plt.scatter(X_test, pred_x, label='predict')
plt.legend()
plt.show()
def test_simple_grnn(self):
dataset = datasets.load_diabetes()
x_train, x_test, y_train, y_test = train_test_split(
dataset.data, dataset.target, train_size=0.7,
random_state=0
)
nw = GRNN(standard_deviation=0.1)
nw.train(x_train, y_train)
result = nw.predict(x_test)
error = rmsle(result, y_test)
self.assertAlmostEqual(error, 0.4245, places=4)
df_cluster = X_train
df_cluster[
'Grade'] = Y_train #Reproduce original data but only with training values
rbfk_net = RBFKMeans(
n_clusters=prototypes) #Chose number of clusters that you want
rbfk_net.train(df_cluster, epsilon=1e-5)
center = pd.DataFrame(rbfk_net.centers)
# Turn the centers into prototypes values needed
X_prototypes = center.iloc[:, 0:-1]
Y_prototypes = center.iloc[:,
-1] #Y_prototypes is the last column of center since 'Grade' is the last feature added to center.
#Train GRNN
GRNNet = GRNN(std=0.1)
GRNNet.train(X_prototypes, Y_prototypes)
# Cross validataion
score = cross_val_score(GRNNet, X_train, Y_train, scoring='r2', cv=5)
print("")
print("Cross Validation: {0} (+/- {1})".format(score.mean().round(2),
(score.std() * 2).round(2)))
print("")
#Prediction
Y_predict = GRNNet.predict(X)
print(Y.values * minmax + minval)
print((Y_predict * minmax + minval)[:, 0].round(2))
print("")
print("Accuracy: {0}".format(metrics.r2_score(Y, Y_predict).round(2)))
print("")