def load_binary_syntetic(dataset, n_train):
# splitting syntetic dataset
X, Y = load_syntetic_dataset(dataset)
Xnew = []
Ynew = []
for x, y in zip(X, Y):
Xnew.append((x[0], x[1], np.ones((x[2].shape[0], 1))))
y_ = y[:, 0].astype(np.int32)
labels = np.unique(y_)
y[y_ == labels[0], 0] = 0
for l in labels[1:]:
if l == 0:
continue
y[y_ == l, 0] = 1
Ynew.append(y)
X = Xnew
Y = Ynew
x_train = X[:n_train]
y_train = [Label(y[:, 0].astype(np.int32), None, y[:, 1], True)
for y in Y[:n_train]]
x_test = X[n_train:]
y_test = [Label(y[:, 0].astype(np.int32), None, y[:, 1], True)
for y in Y[n_train:]]
return x_train, y_train, x_test, y_test
def load_binary_syntetic(dataset, n_train):
# splitting syntetic dataset
X, Y = load_syntetic_dataset(dataset)
Xnew = []
Ynew = []
for x, y in zip(X, Y):
Xnew.append((x[0], x[1], np.ones((x[2].shape[0], 1))))
y_ = y[:, 0].astype(np.int32)
labels = np.unique(y_)
y[y_ == labels[0], 0] = 0
for l in labels[1:]:
if l == 0:
continue
y[y_ == l, 0] = 1
Ynew.append(y)
X = Xnew
Y = Ynew
x_train = X[:n_train]
y_train = [
Label(y[:, 0].astype(np.int32), None, y[:, 1], True)
for y in Y[:n_train]
]
x_test = X[n_train:]
y_test = [
Label(y[:, 0].astype(np.int32), None, y[:, 1], True)
for y in Y[n_train:]
]
return x_train, y_train, x_test, y_test
def load_syntetic(dataset, n_full, n_train):
# splitting syntetic dataset
X, Y = load_syntetic_dataset(dataset)
x_train = X[:n_train]
y_train = [Label(y[:, 0].astype(np.int32), None, y[:, 1], True)
for y in Y[:n_full]]
y_train += [Label(None, np.unique(y[:, 0].astype(np.int32)), y[:, 1], False)
for y in Y[(n_full):(n_train)]]
y_train_full = [Label(y[:, 0].astype(np.int32), None, y[:, 1], True)
for y in Y[:n_train]]
x_test = X[n_train:]
y_test = [Label(y[:, 0].astype(np.int32), None, y[:, 1], True)
for y in Y[n_train:]]
return x_train, y_train, y_train_full, x_test, y_test
def load_syntetic(dataset, n_full, n_train):
# splitting syntetic dataset
X, Y = load_syntetic_dataset(dataset)
x_train = X[:n_train]
y_train = [
Label(y[:, 0].astype(np.int32), None, y[:, 1], True)
for y in Y[:n_full]
]
y_train += [
Label(None, np.unique(y[:, 0].astype(np.int32)), y[:, 1], False)
for y in Y[(n_full):(n_train)]
]
y_train_full = [
Label(y[:, 0].astype(np.int32), None, y[:, 1], True)
for y in Y[:n_train]
]
x_test = X[n_train:]
y_test = [
Label(y[:, 0].astype(np.int32), None, y[:, 1], True)
for y in Y[n_train:]
]
return x_train, y_train, y_train_full, x_test, y_test
