idq.stream

describe module, architecture, etc.

class idq.stream.StreamProcessor(start, end, buffer_factory, buffer_kwargs={}, **manager_kwargs)

handles data processing logic for the streaming pipeline

configure_segdb(seg_config, reporter=None)

we have to configure_segdb in order to not ignore_segdb, and we assume that one would only configure_segdb if you wanted to use it thus, we manage the boolean flag controlling whether to use segdb based on whether this method was called

poll(**triggers_kwargs)

the workhorse method manages requests for data and I/O via ClassifierData objects constructs and returns an umbrella of the requested ClassifierDatas for this stride

triggers_kwargs is passed to calls to ClassifierData.triggers and is exposed to help control I/O costs

query_segdb(start, end)

a helper function that manages segment construction for each buffer

wait()

delegates to CadenceManager.wait(), which updates timestamp after sleeping

idq.stream.calibrate(config_path, gps_start=None, gps_end=None)

calibrate the classifier predictions based on historical data

idq.stream.evaluate(config_path, gps_start=None, gps_end=None)

the realtime evaluation routine

idq.stream.gps_range(gps_start=None, gps_end=None)

figure out the GPS range for a streaming analysis

idq.stream.report(config_path, gps_start=None, gps_end=None, reportdir=None)

run batch.report jobs periodically

idq.stream.stream(config_path, gps_start=None, gps_end=None, workflow='fork', initial_lookback=3600, monitoring_cadence=1.0, skip_timeseries=False, skip_report=False, persist=False, verbose=False, quiet=False)

launch and manage all streaming processes

idq.stream.timeseries(config_path, gps_start=None, gps_end=None)

the realtime evaluation routine

idq.stream.train(config_path, gps_start=None, gps_end=None, initial_lookback=3600)

run the training process