def mnist_regular_graph(k=8):
''' load mnist dataset as graph. The graph is created using knn.
You can change the number of neighbors per node by changing k.
params:
k: number of neighbors per node
returns:
X_train, y_train, X_val, y_val, X_test, y_test, A, node_positions
'''
X_train, y_train, X_val, y_val, X_test, y_test, A = mnist.load_data(k)
X_train = X_train[..., np.newaxis]
X_test = X_test[..., np.newaxis]
y_train = keras.utils.to_categorical(y_train, 10)
y_test = keras.utils.to_categorical(y_test, 10)
n_x, n_y = (28, 28)
x = np.linspace(0, 10, n_x)
y = np.linspace(0, 10, n_y)
y = np.flip(y)
xv, yv = np.meshgrid(x, y)
pos = np.stack([xv.flatten(), yv.flatten()], axis=-1)
return X_train, y_train, X_val, y_val, X_test, y_test, A.toarray(), pos
def test_mnist():
mnist.load_data()
from keras.regularizers import l2
from spektral.datasets import mnist
from spektral.layers import GraphConv
from spektral.layers.ops import sp_matrix_to_sp_tensor
from spektral.utils import normalized_laplacian
# Parameters
l2_reg = 5e-4 # Regularization rate for l2
learning_rate = 1e-3 # Learning rate for SGD
batch_size = 32 # Batch size
epochs = 20000 # Number of training epochs
es_patience = 200 # Patience fot early stopping
# Load data
X_train, y_train, X_val, y_val, X_test, y_test, adj = mnist.load_data()
X_train, X_val, X_test = X_train[..., None], X_val[..., None], X_test[..., None]
N = X_train.shape[-2] # Number of nodes in the graphs
F = X_train.shape[-1] # Node features dimensionality
n_out = y_train.shape[-1] # Dimension of the target
fltr = normalized_laplacian(adj)
# Model definition
X_in = Input(shape=(N, F))
# Pass A as a fixed tensor, otherwise Keras will complain about inputs of
# different rank.
A_in = Input(tensor=sp_matrix_to_sp_tensor(fltr))
graph_conv = GraphConv(32,
activation='elu',
def test_mnist():
mnist.load_data(k=8, noise_level=0.1)
from tensorflow.keras.regularizers import l2
from spektral.datasets import mnist
from spektral.layers import GraphConv
from spektral.layers.ops import sp_matrix_to_sp_tensor
from spektral.utils import batch_iterator
# Parameters
learning_rate = 1e-3 # Learning rate for Adam
batch_size = 32 # Batch size
epochs = 1000 # Number of training epochs
patience = 10 # Patience for early stopping
l2_reg = 5e-4 # Regularization rate for l2
# Load data
x_tr, y_tr, x_va, y_va, x_te, y_te, A = mnist.load_data()
x_tr, x_va, x_te = x_tr[..., None], x_va[..., None], x_te[..., None]
N = x_tr.shape[-2] # Number of nodes in the graphs
F = x_tr.shape[-1] # Node features dimensionality
n_out = 10 # Dimension of the target
# Create filter for GCN and convert to sparse tensor
fltr = GraphConv.preprocess(A)
fltr = sp_matrix_to_sp_tensor(fltr)
# Build model
class Net(Model):
def __init__(self, **kwargs):
super().__init__(**kwargs)
self.conv1 = GraphConv(32, activation='elu', kernel_regularizer=l2(l2_reg))