def yield_datasets(model): from creme import base from creme import datasets from creme import stream from sklearn import datasets as sk_datasets model = guess_model(model) if isinstance(model, (base.BinaryClassifier, base.MultiClassifier)): yield datasets.Phishing() if isinstance(model, base.MultiClassifier): yield datasets.ImageSegments().take(500) if isinstance(model, base.Regressor): yield datasets.TrumpApproval() if isinstance(model, base.MultiOutputRegressor): yield stream.iter_sklearn_dataset(sk_datasets.load_linnerud()) if isinstance(model, base.MultiOutputClassifier): yeast = stream.iter_sklearn_dataset(sk_datasets.fetch_openml('yeast', version=4)) yield itertools.islice(yeast, 100)
def yield_datasets(model): from creme import base from creme import datasets from creme import stream from sklearn import datasets as sk_datasets model = guess_model(model) if isinstance(model, (base.BinaryClassifier, base.MultiClassifier)): yield datasets.Phishing() if isinstance(model, base.MultiClassifier): yield datasets.ImageSegments().take(500) if isinstance(model, base.Regressor): yield datasets.TrumpApproval() if isinstance(model, base.MultiOutputRegressor): yield stream.iter_sklearn_dataset(sk_datasets.load_linnerud()) if isinstance(model, base.MultiOutputClassifier): yield datasets.Music()
def yield_datasets(model): from creme import base from creme import compose from creme import datasets from creme import preprocessing from creme import stream from sklearn import datasets as sk_datasets model = guess_model(model) # Classification if isinstance(model, (base.BinaryClassifier, base.MultiClassifier)): yield datasets.Phishing() # Multi-class classification if isinstance(model, base.MultiClassifier): yield datasets.ImageSegments().take(500) # Regression if isinstance(model, base.Regressor): yield datasets.TrumpApproval() # Multi-output regression if isinstance(model, base.MultiOutputRegressor): # 1 yield stream.iter_sklearn_dataset(sk_datasets.load_linnerud()) # 2 class SolarFlare: """One-hot encoded version of `datasets.SolarFlare`.""" def __iter__(self): oh = (compose.SelectType(str) | preprocessing.OneHotEncoder()) + compose.SelectType(int) for x, y in datasets.SolarFlare(): yield oh.transform_one(x), y yield SolarFlare() # Multi-output classification if isinstance(model, base.MultiOutputClassifier): yield datasets.Music()
def yield_datasets(model): from creme import compose from creme import datasets from creme import preprocessing from creme import stream from creme import utils from sklearn import datasets as sk_datasets # Classification if utils.inspect.isclassifier(model): yield datasets.Phishing() # Multi-class classification if model._multiclass: yield datasets.ImageSegments().take(500) # Regression if utils.inspect.isregressor(model): yield datasets.TrumpApproval() # Multi-output regression if utils.inspect.ismoregressor(model): # 1 yield stream.iter_sklearn_dataset(sk_datasets.load_linnerud()) # 2 class SolarFlare: """One-hot encoded version of `datasets.SolarFlare`.""" def __iter__(self): oh = (compose.SelectType(str) | preprocessing.OneHotEncoder()) + compose.SelectType(int) for x, y in datasets.SolarFlare(): yield oh.transform_one(x), y yield SolarFlare() # Multi-output classification if utils.inspect.ismoclassifier(model): yield datasets.Music()
functools.partial(torch.optim.Adagrad, lr=LR), optimizers.Adagrad(lr=LR)), 'Momentum': (optim.Momentum(lr=LR, rho=.1), functools.partial(torch.optim.SGD, lr=LR, momentum=.1), optimizers.SGD(lr=LR, momentum=.1)) } def add_intercept(x): return {**x, 'intercept': 1.} for name, (creme_optim, torch_optim, keras_optim) in OPTIMIZERS.items(): X_y = stream.iter_sklearn_dataset(dataset=datasets.load_boston(), shuffle=True, random_state=42) n_features = 13 creme_lin_reg = (compose.FuncTransformer(add_intercept) | linear_model.LinearRegression( optimizer=creme_optim, l2=0, intercept_lr=0)) torch_model = PyTorchNet(n_features=n_features) torch_lin_reg = PyTorchRegressor(network=torch_model, loss_fn=torch.nn.MSELoss(), optimizer=torch_optim( torch_model.parameters())) inputs = layers.Input(shape=(n_features, )) predictions = layers.Dense(1,