def __init__(self,
leaf_model=LinearSVC(),
split_classifier=LinearSVC(),
num_features_per_node=None,
max_depth=3,
min_leaf_size=50,
randomize_split_params={},
randomize_leaf_params={},
verbose=False):
# check everyone's types -- I can't give up the OCaml instincts
# also, if running this code remotely it's nice to know when something
# goes wrong before we send an object over to AWS
check_estimator(leaf_model)
check_estimator(split_classifier)
check_int(max_depth)
check_int(min_leaf_size)
check_dict(randomize_split_params)
check_dict(randomize_leaf_params)
check_bool(verbose)
self.leaf_model = leaf_model
self.split_classifier = split_classifier
self.max_depth = max_depth
self.min_leaf_size = min_leaf_size
self.num_features_per_node = num_features_per_node
self.randomize_split_params = randomize_split_params
self.randomize_leaf_params = randomize_leaf_params
self.verbose = verbose
self.root = None
self.classes = None
def __init__(self,
leaf_model = LinearSVC(),
split_classifier = LinearSVC(),
num_features_per_node = None,
max_depth=3,
min_leaf_size=50,
randomize_split_params={},
randomize_leaf_params={},
verbose = False):
# check everyone's types -- I can't give up the OCaml instincts
# also, if running this code remotely it's nice to know when something
# goes wrong before we send an object over to AWS
check_estimator(leaf_model)
check_estimator(split_classifier)
check_int(max_depth)
check_int(min_leaf_size)
check_dict(randomize_split_params)
check_dict(randomize_leaf_params)
check_bool(verbose)
self.leaf_model = leaf_model
self.split_classifier = split_classifier
self.max_depth = max_depth
self.min_leaf_size = min_leaf_size
self.num_features_per_node = num_features_per_node
self.randomize_split_params = randomize_split_params
self.randomize_leaf_params = randomize_leaf_params
self.verbose = verbose
self.root = None
self.classes = None
def __init__(self,
base_model,
num_models,
bagging_percent,
bagging_replacement,
feature_subset_percent,
stacking_model,
randomize_params,
additive,
verbose):
check_estimator(base_model)
check_int(num_models)
self.base_model = base_model
self.num_models = num_models
self.bagging_percent = bagging_percent
self.bagging_replacement = bagging_replacement
self.feature_subset_percent = feature_subset_percent
self.stacking_model = stacking_model
self.randomize_params = randomize_params
self.additive = additive
self.verbose = verbose
self.need_to_fit = True
self.models = None
self.weights = None
def __init__(self, k, base_model, verbose=False):
check_int(k)
check_estimator(base_model)
check_bool(verbose)
self.k = k
self.base_model = base_model
self.verbose = verbose
self.clusters = MiniBatchKMeans(k)
self.models = None