def __init__(self, uuid, **kwargs):
# Call super class init first
model.__init__(self, uuid, **kwargs)
self._feature_vector_paths = kwargs.get("feature_vectors")
self._truth_vector_paths = kwargs.get("truth_vectors")
# Handle to storage for model parameters
params = self._parameters
# Set various bits of meta data, these are defaults and can be changed later
params.setMetaData("model_type", u"decision_tree_regressor_model")
params.setMetaData("model_desc",
u"SciKits Decision Tree Regressor implementation")
params.setMetaData("criterion", kwargs.get("criterion"))
params.setMetaData("splitter", kwargs.get("splitter"))
params.setMetaData("max_features", kwargs.get("max_features"))
params.setMetaData("max_depth", kwargs.get("max_depth"))
params.setMetaData("min_samples_split",
kwargs.get("min_samples_split"))
params.setMetaData("min_samples_leaf", kwargs.get("min_samples_leaf"))
params.setMetaData("max_weight_fraction_leaf",
kwargs.get("max_weight_fraction_leaf"))
params.setMetaData("max_leaf_nodes", kwargs.get("max_leaf_nodes"))
params.setMetaData("random_state", kwargs.get("random_state"))
params.setMetaData("presort", kwargs.get("presort"))
# Validation fields
# params.addRequiredBinaryFields([
# "lr_coef_json",
# "lr_intercept_json"])
# Validator information
params.setMetaData("db_views", [])
def __init__(self, uuid, **kwargs):
# Call super class init first
model.__init__(self, uuid, **kwargs)
gamma = kwargs.get("gamma")
if (gamma is None):
gamma = 0.001
# Handle to storage for model parameters
# Set various bits of meta data, these are defaults and can be changed later
self._parameters.setMetaData("model_type", u"example_model")
self._parameters.setMetaData("model_desc", u"A description of the example_model")
self._parameters.setMetaData("gamma", gamma)
self._parameters.setMetaData("clf", {})
self._parameters.setMetaData("feature_vectors", [])
self._parameters.setMetaData("truth_vectors", [])
def __init__(self, uuid, **kwargs):
# Call super class init first
model.__init__(self, uuid, **kwargs)
self._feature_vector_paths = kwargs.get("feature_vectors")
self._truth_vector_paths = kwargs.get("truth_vectors")
# Handle to storage for model parameters
params = self._parameters
# Set various bits of meta data, these are defaults and can be changed later
params.setMetaData("model_type", u"gaussian_nb_model")
params.setMetaData("model_desc",
u"SciKits Gaussian Naive Bayes implementation")
# Validator information
params.setMetaData("db_views", [])
def __init__(self, uuid, **kwargs):
# Call super class init first
model.__init__(self, uuid, **kwargs)
# Handle to storage for model parameters
params = self._parameters
# Set various bits of meta data, these are defaults and can be changed later
params.setMetaData("model_type", u"multinomial_nb_model")
params.setMetaData("model_desc",
u"SciKits Multinomial Naive Bayes implementation")
params.setMetaData("alpha", kwargs.get("alpha"))
params.setMetaData("fit_prior", kwargs.get("fit_prior"))
params.setMetaData("class_prior", kwargs.get("class_prior"))
# Validator information
params.setMetaData("db_views", [])
def __init__(self, uuid, **kwargs):
""" Create instance of the db management model
Input:
kwargs: various parameters
"""
# Call super class init first
model.__init__(self, uuid, **kwargs)
# Handle to storage for model parameters
params = self._parameters
# Set various bits of meta data, these are defaults and can be changed later
params.setMetaData("model_type", u"attribute_index")
params.setMetaData("model_desc", u"The most awesome attribute_index model ever")
params.setMetaData("value_index", {})
params.setMetaData("processed_count", 0)
params.setMetaData("count_threshold", 0)
attributes_to_index = kwargs.get("attributes_to_index")
if attributes_to_index is None:
attributes_to_index = []
values_to_index = kwargs.get("values_to_index")
if values_to_index is None:
values_to_index = []
count_threshold = kwargs.get("count_threshold")
if (count_threshold == None):
count_threshold = 0
params.setMetaData("attributes_to_index", attributes_to_index)
params.setMetaData("values_to_index", values_to_index)
params.setMetaData("count_threshold", count_threshold)
self._hyperparameter_list.append("attributes_to_index")
self._hyperparameter_list.append("values_to_index")
self._hyperparameter_list.append("count_threshold")
def __init__(self, uuid, **kwargs):
# Call super class init first
print "lr_model kwargs " + str(kwargs)
model.__init__(self, uuid, **kwargs)
self._feature_vector_paths = kwargs.get("feature_vectors")
self._truth_vector_paths = kwargs.get("truth_vectors")
# Handle to storage for model parameters
params = self._parameters
# Set various bits of meta data, these are defaults and can be changed later
params.setMetaData("model_type", u"logistic_regression_model")
params.setMetaData("model_desc",
u"SciKits Logistic Regression implementation")
params.setMetaData("penalty", kwargs.get("penalty"))
params.setMetaData("dual", kwargs.get("dual"))
params.setMetaData("C", kwargs.get("C"))
params.setMetaData("fit_intercept", kwargs.get("fit_intercept"))
params.setMetaData("intercept_scaling",
kwargs.get("intercept_scaling"))
params.setMetaData("class_weight", kwargs.get("class_weight"))
params.setMetaData("max_iter", kwargs.get("max_iter"))
params.setMetaData("random_state", kwargs.get("random_state"))
params.setMetaData("solver", kwargs.get("solver"))
params.setMetaData("tol", kwargs.get("tol"))
params.setMetaData("multi_class", kwargs.get("multi_class"))
params.setMetaData("verbose", kwargs.get("verbose"))
# Validation fields
# params.addRequiredBinaryFields([
# "lr_coef_json",
# "lr_intercept_json"])
# Validator information
params.setMetaData("db_views", [])
def __init__(self, uuid, **kwargs):
""" Create instance of the db management model
Input:
kwargs: various parameters
"""
# Call super class init first
model.__init__(self, uuid, **kwargs)
# Handle to storage for model parameters
params = self._parameters
# Set various bits of meta data, these are defaults and can be changed later
params.setMetaData("model_type", u"attribute_count_analysis")
params.setMetaData(
"model_desc",
u"The most awesome attribute_count_analysis model ever")
params.setMetaData("attribute_index_name",
kwargs.get("attribute_index_name"))
params.setMetaData("attribute_index_db_type",
kwargs.get("attribute_index_db_type"))
params.setMetaData("attribute_index_db_host",
kwargs.get("attribute_index_db_host"))
params.setMetaData("attribute_index_db_name",
kwargs.get("attribute_index_db_name"))
antecedent_key_list = kwargs.get("antecedent_key_list")
if antecedent_key_list is None:
antecedent_key_list = []
antecedent_value_list = kwargs.get("antecedent_value_list")
if antecedent_value_list is None:
antecedent_value_list = []
consequent_key_list = kwargs.get("consequent_key_list")
if consequent_key_list is None:
consequent_key_list = []
consequent_value_list = kwargs.get("consequent_value_list")
if consequent_value_list is None:
consequent_value_list = []
params.setMetaData("antecedent_key_list", antecedent_key_list)
params.setMetaData("antecedent_value_list", antecedent_value_list)
params.setMetaData("consequent_key_list", consequent_key_list)
params.setMetaData("consequent_value_list", consequent_value_list)
# report_antecedent_count = kwargs.get("report_antecedent_count")
# if (report_antecedent_count==None):
# report_antecedent_count = False
report_tp = kwargs.get("report_tp")
if report_tp is None:
report_tp = False
report_fp = kwargs.get("report_fp")
if report_fp is None:
report_fp = False
report_fn = kwargs.get("report_fn")
if report_fn is None:
report_fn = False
report_precision = kwargs.get("report_precision")
if report_precision is None:
report_precision = False
report_recall = kwargs.get("report_recall")
if report_recall is None:
report_recall = False
report_f_measure = kwargs.get("report_f_measure")
if report_f_measure is None:
report_f_measure = False
# params.setMetaData("report_antecedent_count", report_antecedent_count)
params.setMetaData("report_tp", report_tp)
params.setMetaData("report_fp", report_fp)
params.setMetaData("report_fn", report_fn)
params.setMetaData("report_precision", report_precision)
params.setMetaData("report_recall", report_recall)
params.setMetaData("report_f_measure", report_f_measure)
self._hyperparameter_list.append("attribute_index_name")
self._hyperparameter_list.append("attribute_index_db_type")
self._hyperparameter_list.append("attribute_index_db_host")
self._hyperparameter_list.append("attribute_index_db_name")
self._hyperparameter_list.append("antecedent_key_list")
self._hyperparameter_list.append("antecedent_value_list")
self._hyperparameter_list.append("consequent_key_list")
self._hyperparameter_list.append("consequent_value_list")
self._hyperparameter_list.append("report_tp")
self._hyperparameter_list.append("report_fp")
self._hyperparameter_list.append("report_fn")
self._hyperparameter_list.append("report_precision")
self._hyperparameter_list.append("report_recall")
self._hyperparameter_list.append("report_f_measure")
params.setMetaData("value_lists", {})
params.setMetaData("tp_lists", {})
params.setMetaData("fp_lists", {})
params.setMetaData("fn_lists", {})
params.setMetaData("stats_dicts", {})