def __full_init__(self,
X,
Y,
name=None,
pipeline=None,
K=5,
kernel="euclidean",
algo="auto",
weights="uniform",
kernel_params={},
client=None):
self.X = X
self.Y = Y
if name == None:
name = rando_name()
self.name = name
self.pipeline = pipeline
self.K = K
self.kernel = kernel
self.kernel_params = kernel_params
self.algo = algo
self.weights = weights
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = KNNRegressorPipeline(pipeline_name, K, kernel,
algo, weights, kernel_params,
client)
self.response = self.pipeline.run(X=self.X, Y=self.Y, model=self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self,
X,
Y,
name=None,
pipeline=None,
n_trees=8,
client=None):
self.n_trees = n_trees
if name == None:
name = rando_name()
# else:
# todo: add feature to instantiate RFRegressor just from name
# i.e. an already created model
self.name = name
self.X = X
self.Y = Y
self.pipeline = pipeline
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = RFRegressorPipeline(pipeline_name, n_trees, client)
self.response = self.pipeline.run(X=self.X, Y=self.Y, model=self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self,
dataset,
name=None,
pipeline=None,
init_alpha=None,
init_beta=None,
trans_cov=None,
obs_cov=None,
init_cov=None,
optimizations=[],
client=None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.init_alpha = init_alpha
self.init_beta = init_beta
self.trans_cov = trans_cov
self.obs_cov = obs_cov
self.init_cov = init_cov
self.optimizations = optimizations
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = KalmanOLSPipeline(pipeline_name, init_alpha,
init_beta, trans_cov, obs_cov,
init_cov, optimizations, client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self,
dataset,
class_column,
name=None,
pipeline=None,
k=5,
kernel="euclidean",
algo="auto",
weights="uniform",
kernel_params={},
client=None):
self.class_column = class_column
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = KNNClassifierPipeline(pipeline_name, k, kernel,
algo, weights, kernel_params,
client)
self.response = self.pipeline.run(self.dataset, self.name,
self.class_column)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self,
dataset,
name=None,
pipeline=None,
K=3,
kernel="euclidean",
client=None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.K = K
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = NeighborNetworkGraphPipeline(
pipeline_name, K, kernel, client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self, dataset, name = None, pipeline = None, K = 5, kernel = "euclidean", geodesic = False, kernel_params = {}, client = None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.kernel = kernel
self.kernel_params = kernel_params
self.K = K
self.geodesic = geodesic
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = DistanceMatrixPipeline(name = pipeline_name,
K = K,
kernel = kernel,
geodesic = geodesic,
kernel_params = kernel_params,
client = client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self,
label_dataset,
matrix=None,
matrix_name=None,
matrix_type=None,
name=None,
pipeline=None,
alpha=0.1,
client=None):
if name == None:
name = rando_name()
self.name = name
self.label_dataset = label_dataset
self.pipeline = pipeline
self.matrix = matrix
self.matrix_name = matrix_name
self.matrix_type = matrix_type
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = MatrixAgglomeratorPipeline(name=pipeline_name,
alpha=alpha,
client=client)
self.response = self.pipeline.run(label_dataset=self.label_dataset,
model=self.name,
matrix=matrix,
matrix_name=matrix_name,
matrix_type=matrix_type)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self,
X,
Y,
name=None,
pipeline=None,
kernel="linear",
alpha=1.0,
kernel_params={},
client=None):
self.X = X
self.Y = Y
if name == None:
name = rando_name()
self.name = name
self.pipeline = pipeline
self.kernel = kernel
self.alpha = alpha
self.kernel_params = kernel_params
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = KernelRidgeRegressorPipeline(
pipeline_name, kernel, alpha, kernel_params, client)
self.response = self.pipeline.run(X=self.X, Y=self.Y, model=self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self,
matrix,
matrix_name=None,
matrix_type=None,
name=None,
pipeline=None,
client=None):
if name == None:
name = rando_name()
self.name = name
self.matrix = matrix
self.pipeline = pipeline
self.matrix_name = matrix_name
self.matrix_type = matrix_type
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = MatrixMinimumSpanningTreePipeline(
name=pipeline_name, client=client)
self.response = self.pipeline.run(model=self.name,
matrix=self.matrix,
matrix_name=self.matrix_name,
matrix_type=self.matrix_type)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self, dataset, name = None, pipeline = None, retrain = True, client = None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.retrain = retrain
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = BasicA2DPipeline(pipeline_name, retrain, client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self, dataset, name = None, pipeline = None, bandwidth = "scott", client = None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.bandwidth = bandwidth
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = KernelDensityEstimatorPipeline(pipeline_name, bandwidth, client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self, dataset, name = None, pipeline = None, rate = 0.1, n_trees = 100, client = None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.rate = rate
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = IsolationForestPipeline(pipeline_name, rate, n_trees, client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self, dataset, name = None, pipeline = None, D = 2, affinity = "knn", K = 5, gamma = 1.0, client = None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.D = D
self.K = K
self.affinity = affinity
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = LaplacianEigenmapperPipeline(pipeline_name, D, affinity, K, gamma, client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self, dataset, name = None, pipeline = None, D = 2, k = 3, method = "standard", client = None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.D = D
self.k = k
self.method = method
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = LocalLinearEmbedderPipeline(pipeline_name, D, k, method, client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self, dataset, name = None, pipeline = None, D = 2, K = 3, client = None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.D = D
self.K = K
self.type = "raw_isomap"
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = IsomapPipeline(pipeline_name, D, K, client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self, dataset, name = None, pipeline = None, D = 2, kernel = "linear", alpha = 1.0, invert = False, kernel_params = {}, client = None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.D = D
self.kernel = kernel
self.alpha = alpha
self.invert = invert
self.kernel_params = kernel_params
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = KernelPCAPipeline(pipeline_name, D, kernel, alpha, invert, kernel_params, client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self,
dataset,
name=None,
pipeline=None,
sigma=0.1,
client=None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = PNNClassifierPipeline(pipeline_name, sigma, client)
self.response = self.pipeline.run(self.dataset, self.name,
self.class_column)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)
def __full_init__(self,
dataset,
name=None,
pipeline=None,
corr_method="pearson",
client=None):
if name == None:
name = rando_name()
self.name = name
self.dataset = dataset
self.pipeline = pipeline
self.corr_method = corr_method
if self.pipeline == None:
pipeline_name = rando_name()
self.pipeline = MinimumSpanningTreePipeline(
pipeline_name, corr_method, client)
self.response = self.pipeline.run(self.dataset, self.name)
try:
# model will be key if success
model = self.response['model']
self.name = model.split("/")[-1]
except:
# something went wrong creating the model
raise StandardError(self.response)