Classifying A Table of Events

Introduction

Gravity Spy is designed to generally classify a while table of excess noise events at once. These events are then generally, although not always, uploaded to the Zooniverse website. In addition, these events are assumed to have originated from the Event Trigger Generator Omicron. With this in mind, the Events class provides a host of methods that server as wrappers on tables of events to help with a varity of tasks described below.

A Table of Events

doing stuff

In [1]: from gwpy.timeseries import TimeSeries

In [2]: from gravityspy.table import Events

In [3]: timeseries = TimeSeries.read('data/timeseries/L-L1_SOFTWAREINJ-1173197648-132.h5')

In [4]: triggers = Events.read('data/omicron/L1-DCH_FAKE_STRAIN_16k_BBH-SEOBNRv3_OMICRON-1173188120-14996.xml.gz', format='ligolw')

In [5]: results = triggers.classify(path_to_cnn='../models/multi_view_classifier.h5',
   ...:                             timeseries=timeseries, nproc=3)
   ...: 

In [6]: print(results)
 1080Lines  1400Ripples Air_Compressor ... url4   event_time     q_value   
----------- ----------- -------------- ... ---- ------------- -------------
0.000691219 1.40814e-06     0.00771667 ...      1173197717.33 45.2548339959
 0.00475575 5.35774e-05       0.762743 ...      1173197716.53 45.2548339959
 0.00202278 1.35975e-06     0.00132722 ...      1173197715.31 45.2548339959