def generateAlertsCsvFile(self, directory):
detection_threshold = self.alerts_conf.detection_threshold
with open(directory + 'alerts.csv', 'w') as f:
alerts = matrix_tools.extractRowsWithThresholds(
self.predictions_monitoring.predictions, detection_threshold,
None, 'predicted_proba')
alerts = matrix_tools.sortDataFrame(alerts, 'predicted_proba',
False, False)
alerts.to_csv(f, index_label='instance_id')
return list(alerts.index.values)
def getPredictions(experiment_id, train_test, index):
experiment = updateCurrentExperiment(experiment_id)
filename = experiment.getOutputDirectory()
filename += train_test + '/predictions.csv'
index = int(index)
min_value = index * 0.1
max_value = (index + 1) * 0.1
with open(filename, 'r') as f:
data = pd.read_csv(f, header=0, index_col=0)
data = matrix_tools.extractRowsWithThresholds(data, min_value,
max_value,
'predicted_proba')
selected_instances = list(data.index.values)
proba = list(data['predicted_proba'])
return jsonify({'instances': selected_instances, 'proba': proba})
def runModels(self, already_queried = None):
df = matrix_tools.extractRowsWithThresholds(self.predictions,
self.proba_min, self.proba_max, 'predicted_proba')
if already_queried is not None:
self.predicted_ids = list(set(df.index).difference(set(already_queried)))
else:
self.predicted_ids = list(df.index)
datasets = self.iteration.datasets
self.annotated_instances = datasets.getAnnotatedInstances(label = self.label)
self.families_analysis = self.familiesAnalysis()
if self.families_analysis:
self.annotations_type = 'families'
start_time = time.time()
self.buildCategories()
self.analysis_time = time.time() - start_time
self.categories.setLikelihood(self.iteration.iteration_number)
else:
self.annotations_type = 'individual'
self.categories = None
self.analysis_time = 0