def svm_prediction(model, kernel, model_id, predictors):
'''@svm_prediction
This method generates an svm prediction using the provided prediction
feature input(s), and the stored corresponding model, within the NoSQL
datastore.
@predictors, a list of arguments (floats) required to make an SVM
prediction, against the respective svm model.
'''
# get necessary model
title = Cache_Hset().uncache(
model + '_' + kernel + '_title',
model_id
)['result']
clf = Cache_Model().uncache(
model + '_' + kernel + '_model',
model_id + '_' + title
)
# get encoded labels
encoded_labels = Cache_Model().uncache(
model + '_' + kernel + '_labels',
model_id
)
# perform prediction, and return the result
numeric_label = (clf.predict([predictors]))
textual_label = list(encoded_labels.inverse_transform([numeric_label]))
return {'result': textual_label[0][0], 'error': None}
def sv_prediction(model, model_id, predictors):
'''@sv_prediction
This method generates an sv (i.e. svm, or svr) prediction using the
provided prediction feature input(s), and the stored corresponding model,
within the NoSQL datastore.
@clf, decoded model, containing several methods (i.e. predict)
@predictors, a list of arguments (floats) required to make an SVM
prediction, against the respective svm model.
'''
# local variables
list_model_type = current_app.config.get('MODEL_TYPE')
# get necessary model
title = Cache_Hset().uncache(
model + '_title',
model_id
)['result']
clf = Cache_Model().uncache(
model + '_model',
model_id + '_' + title
)
# svm model: get encoded labels
if model == list_model_type[0]:
encoded_labels = Cache_Model().uncache(
model + '_labels',
model_id
)
# perform prediction
numeric_label = (clf.predict([predictors]))
# result: svm model
if model == list_model_type[0]:
textual_label = list(encoded_labels.inverse_transform([numeric_label]))
return {'result': textual_label[0][0], 'error': None}
# result: svr model
elif model == list_model_type[1]:
return {'result': str(numeric_label[0]), 'error': None}
def retrieve_sv_features():
'''@retrieve_sv_features
This router function retrieves the generalized features properties that can
be expected for any given observation within the supplied dataset.
@label_list, this value will be a json object, since it was originally
cached into redis using 'json.dumps'.
'''
if request.method == 'POST':
label_list = Cache_Hset().uncache('svm_rbf_feature_labels',
request.get_json()['session_id'])
# return all feature labels
if label_list['result']:
return json.dumps(label_list['result'])
else:
return json.dumps({'error': label_list['error']})
def retrieve_sv_features():
'''
This router function retrieves the generalized features properties that can
be expected for any given observation within the supplied dataset.
@label_list, this value will be a json object, since it was originally
cached into redis using 'json.dumps'.
'''
# get model type
model_id = request.get_json()['model_id']
model_type = M_Type().get_model_type(model_id)['result']
# return all feature labels
if request.method == 'POST':
label_list = Cache_Hset().uncache(model_type + '_feature_labels',
model_id)
if label_list['result']:
return json.dumps(label_list['result'])
else:
return json.dumps({'error': label_list['error']})
def sv_prediction(model, model_id, predictors):
'''
This method generates an sv (i.e. svm, or svr) prediction using the
provided prediction feature input(s), and the stored corresponding model,
within the NoSQL datastore.
Additionally, the following is returned for SVM predictions:
- array of probability a given point (predictors) is one of the
defined set of classifiers.
- array of sum distances a given point (predictors) is to the set
of associated hyperplanes.
However, the following is returned for SVR predictions:
- coefficient of determination (r^2).
@clf, decoded model, containing several methods (i.e. predict)
@predictors, a list of arguments (floats) required to make an SVM
prediction, against the respective svm model.
'''
# local variables
probability = None
decision_function = None
list_model_type = current_app.config.get('MODEL_TYPE')
# get necessary model
title = Cache_Hset().uncache(
model + '_title',
model_id
)['result']
clf = Cache_Model().uncache(
model + '_model',
model_id + '_' + title
)
# perform prediction, and return the result
prediction = (clf.predict([predictors]))
# case 1: return svm prediction, and confidence level
if model == list_model_type[0]:
encoded_labels = Cache_Model().uncache(
model + '_labels',
model_id
)
textual_label = encoded_labels.inverse_transform([prediction])
probability = clf.predict_proba(predictors)
decision_function = clf.decision_function(predictors)
classes = [encoded_labels.inverse_transform(x) for x in clf.classes_]
return {
'result': textual_label[0][0],
'model': model,
'confidence': {
'classes': list(classes),
'probability': list(probability[0]),
'decision_function': list(decision_function[0])
},
'error': None
}
# case 2: return svr prediction, and confidence level
elif model == list_model_type[1]:
r2 = Cache_Hset().uncache(
model + '_r2',
model_id
)['result']
return {
'result': str(prediction[0]),
'model': model,
'confidence': {
'score': r2
},
'error': None
}
def generate_model(self):
"""@generate_model
This method generates an svm model, using a chosen dataset from the SQL
database. The resulting model is stored into a NoSQL datastore.
@grouped_features, a matrix of observations, where each nested vector,
or python list, is a collection of features within the containing
observation.
@encoded_labels, observation labels (dependent variable labels),
encoded into a unique integer representation.
"""
# local variables
dataset = self.feature_request.get_dataset(self.session_id)
feature_count = self.feature_request.get_count(self.session_id)
label_encoder = preprocessing.LabelEncoder()
# get dataset
if dataset['error']:
print dataset['error']
self.list_error.append(dataset['error'])
dataset = None
else:
dataset = numpy.asarray(dataset['result'])
# get feature count
if feature_count['error']:
print feature_count['error']
self.list_error.append(feature_count['error'])
feature_count = None
else:
feature_count = feature_count['result'][0][0]
# check dataset integrity, build model
if len(dataset) % feature_count == 0:
features_list = dataset[:, [[0], [2], [1]]]
current_features = []
grouped_features = []
observation_labels = []
feature_labels = []
# group features into observation instances, record labels
for index, feature in enumerate(features_list):
if not (index + 1) % feature_count == 0:
# observation labels
current_features.append(feature[1][0])
# general feature labels in every observation
if not len(feature_labels) == feature_count:
feature_labels.append(feature[2][0])
else:
# general feature labels in every observation
if not len(feature_labels) == feature_count:
feature_labels.append(feature[2][0])
current_features.append(feature[1][0])
grouped_features.append(current_features)
observation_labels.append(feature[0][0])
current_features = []
# convert observation labels to a unique integer representation
label_encoder = preprocessing.LabelEncoder()
label_encoder.fit(dataset[:, 0])
encoded_labels = label_encoder.transform(observation_labels)
# create svm model
clf = svm.SVC()
clf.fit(grouped_features, encoded_labels)
# get svm title, and cache (model, encoded labels, title)
entity = Retrieve_Entity()
title = entity.get_title(self.session_id)['result'][0][0]
Cache_Model(clf).cache('svm_rbf_model',
str(self.session_id) + '_' + title)
Cache_Model(label_encoder).cache('svm_rbf_labels', self.session_id)
Cache_Hset().cache('svm_rbf_title', self.session_id, title)
# cache svm feature labels, with respect to given session id
Cache_Hset().cache('svm_rbf_feature_labels', str(self.session_id),
json.dumps(feature_labels))
def sv_model(model, kernel_type, session_id, feature_request, list_error):
'''@sv_model
This method generates an sv (i.e. svm, or svr) model using feature data,
retrieved from the database. The generated model, is then stored within the
NoSQL datastore.
@grouped_features, a matrix of observations, where each nested vector,
or python list, is a collection of features within the containing
observation.
@encoded_labels, observation labels (dependent variable labels),
encoded into a unique integer representation.
'''
# local variables
dataset = feature_request.get_dataset(session_id, model)
get_feature_count = feature_request.get_count(session_id)
label_encoder = preprocessing.LabelEncoder()
logger = Logger(__name__, 'error', 'error')
list_model_type = current_app.config.get('MODEL_TYPE')
# get dataset
if dataset['error']:
logger.log(dataset['error'])
list_error.append(dataset['error'])
dataset = None
else:
dataset = numpy.asarray(dataset['result'])
# get feature count
if get_feature_count['error']:
logger.log(get_feature_count['error'])
list_error.append(get_feature_count['error'])
feature_count = None
else:
feature_count = get_feature_count['result'][0][0]
# check dataset integrity, build model
if len(dataset) % feature_count == 0:
features_list = dataset[:, [[0], [2], [1]]]
current_features = []
grouped_features = []
observation_labels = []
feature_labels = []
# group features into observation instances, record labels
for index, feature in enumerate(features_list):
# svm: observation labels
if model == list_model_type[0]:
current_features.append(feature[1][0])
if (index+1) % feature_count == 0:
grouped_features.append(current_features)
observation_labels.append(feature[0][0])
current_features = []
# svr: observation labels
elif model == list_model_type[1]:
current_features.append(float(feature[1][0]))
if (index+1) % feature_count == 0:
grouped_features.append(current_features)
observation_labels.append(float(feature[0][0]))
current_features = []
# general feature labels in every observation
if not len(feature_labels) == feature_count:
feature_labels.append(feature[2][0])
# case 1: svm model
if model == list_model_type[0]:
# convert observation labels to a unique integer representation
label_encoder = preprocessing.LabelEncoder()
label_encoder.fit(dataset[:, 0])
encoded_labels = label_encoder.transform(observation_labels)
# create model
clf = svm.SVC(kernel=kernel_type, probability=True)
# cache encoded labels
Cache_Model(label_encoder).cache(model + '_labels', session_id)
# fit model
clf.fit(grouped_features, encoded_labels)
# case 2: svr model
elif model == list_model_type[1]:
# create model
clf = svm.SVR(kernel=kernel_type)
# fit model
clf.fit(grouped_features, observation_labels)
# compute, and cache coefficient of determination
r2 = clf.score(grouped_features, observation_labels)
Cache_Hset().cache(
model + '_r2',
session_id,
r2
)
# get title
entity = Retrieve_Entity()
title = entity.get_title(session_id)['result'][0][0]
# cache model, title
Cache_Model(clf).cache(
model + '_model',
str(session_id) + '_' + title
)
Cache_Hset().cache(model + '_title', session_id, title)
# cache feature labels, with respect to given session id
Cache_Hset().cache(
model + '_feature_labels',
str(session_id),
json.dumps(feature_labels)
)
# return error(s) if exists
return {'error': list_error}
def svm_model(kernel_type, session_id, feature_request, list_error):
'''@svm_model
This method generates an svm prediction using the provided prediction
feature input(s), and the stored corresponding model, within the NoSQL
datastore.
@grouped_features, a matrix of observations, where each nested vector,
or python list, is a collection of features within the containing
observation.
@encoded_labels, observation labels (dependent variable labels),
encoded into a unique integer representation.
'''
# local variables
dataset = feature_request.get_dataset(session_id)
feature_count = feature_request.get_count(session_id)
label_encoder = preprocessing.LabelEncoder()
logger = Logger(__name__, 'error', 'error')
# get dataset
if dataset['error']:
logger.log(dataset['error'])
list_error.append(dataset['error'])
dataset = None
else:
dataset = numpy.asarray(dataset['result'])
# get feature count
if feature_count['error']:
logger.log(feature_count['error'])
list_error.append(feature_count['error'])
feature_count = None
else:
feature_count = feature_count['result'][0][0]
# check dataset integrity, build model
if len(dataset) % feature_count == 0:
features_list = dataset[:, [[0], [2], [1]]]
current_features = []
grouped_features = []
observation_labels = []
feature_labels = []
# group features into observation instances, record labels
for index, feature in enumerate(features_list):
if not (index+1) % feature_count == 0:
# observation labels
current_features.append(feature[1][0])
# general feature labels in every observation
if not len(feature_labels) == feature_count:
feature_labels.append(feature[2][0])
else:
# general feature labels in every observation
if not len(feature_labels) == feature_count:
feature_labels.append(feature[2][0])
current_features.append(feature[1][0])
grouped_features.append(current_features)
observation_labels.append(feature[0][0])
current_features = []
# convert observation labels to a unique integer representation
label_encoder = preprocessing.LabelEncoder()
label_encoder.fit(dataset[:, 0])
encoded_labels = label_encoder.transform(observation_labels)
# create svm model
clf = svm.SVC(kernel=kernel_type)
clf.fit(grouped_features, encoded_labels)
# get svm title, and cache (model, encoded labels, title)
entity = Retrieve_Entity()
title = entity.get_title(session_id)['result'][0][0]
Cache_Model(clf).cache(
'svm_rbf_model',
str(session_id) + '_' + title
)
Cache_Model(label_encoder).cache('svm_rbf_labels', session_id)
Cache_Hset().cache('svm_rbf_title', session_id, title)
# cache svm feature labels, with respect to given session id
Cache_Hset().cache(
'svm_rbf_feature_labels',
str(session_id),
json.dumps(feature_labels)
)
# return error(s) if exists
return {'error': list_error}