def _func_create_model(sessionId, **kwargs):
try:
from cuml.neighbors.nearest_neighbors_mg import \
NearestNeighborsMG as cumlNN
except ImportError:
raise_mg_import_exception()
handle = get_raft_comm_state(sessionId)["handle"]
return cumlNN(handle=handle, **kwargs)
def _func_create_model(sessionId, **kwargs):
try:
from cuml.neighbors.kneighbors_classifier_mg import \
KNeighborsClassifierMG as cumlKNN
except ImportError:
raise_mg_import_exception()
handle = get_raft_comm_state(sessionId)["handle"]
return cumlKNN(handle=handle, **kwargs)
def _func_fit(sessionId, objs, datatype, has_weights, **kwargs):
from cuml.cluster.kmeans_mg import KMeansMG as cumlKMeans
handle = get_raft_comm_state(sessionId)["handle"]
if not has_weights:
inp_data = concatenate(objs)
inp_weights = None
else:
inp_data = concatenate([X for X, weights in objs])
inp_weights = concatenate([weights for X, weights in objs])
return cumlKMeans(handle=handle, output_type=datatype,
**kwargs).fit(inp_data,
sample_weight=inp_weights)
def _func(sessionId, data, verbose, **kwargs):
from cuml.cluster.dbscan_mg import DBSCANMG as cumlDBSCAN
handle = get_raft_comm_state(sessionId)["handle"]
return cumlDBSCAN(handle=handle, verbose=verbose,
**kwargs).fit(data, out_dtype=out_dtype)
def _create_model(sessionId, datatype, **kwargs):
from cuml.solvers.cd_mg import CDMG
handle = get_raft_comm_state(sessionId)["handle"]
return CDMG(handle=handle, output_type=datatype, **kwargs)
def _create_model(sessionId, datatype, **kwargs):
from cuml.linear_model.linear_regression_mg import LinearRegressionMG
handle = get_raft_comm_state(sessionId)["handle"]
return LinearRegressionMG(handle=handle, output_type=datatype,
**kwargs)
def _create_model(sessionId, model_func, datatype, **kwargs):
handle = get_raft_comm_state(sessionId)["handle"]
return model_func(handle, datatype, **kwargs)