DataWriter#

class ctapipe.io.DataWriter(**kwargs: Any)[source]#

Bases: Component

Serialize a sequence of events into a HDF5 file, in the correct format

Examples

inside a Tool:

with DataWriter(parent=self) as write_data:
    for event in source:
        calibrate(event)
        process_images(event)
        write_data(event)

Attributes Summary

`compression_level`	compression level, 0=None, 9=maximum
`compression_type`	compressor algorithm to use.
`contact_info`	A trait whose value must be an instance of a specified class.
`context_metadata`	Additional metadata keywords and values that describe this data.
`datalevels`	returns a list of data levels requested
`image_dtype`	A trait for unicode strings.
`image_offset`	An int trait.
`image_scale`	A float trait.
`instrument_info`	A trait whose value must be an instance of a specified class.
`output_path`	output filename
`overwrite`	overwrite output file if it exists
`peak_time_dtype`	A trait for unicode strings.
`peak_time_offset`	An int trait.
`peak_time_scale`	A float trait.
`transform_image`	A boolean (True, False) trait.
`transform_peak_time`	A boolean (True, False) trait.
`transform_waveform`	A boolean (True, False) trait.
`waveform_dtype`	A trait for unicode strings.
`waveform_offset`	An int trait.
`waveform_scale`	A float trait.
`write_dl1_images`	Store DL1 Images if available
`write_dl1_parameters`	Store DL1 image parameters if available
`write_dl2`	Store DL2 stereo shower parameters if available
`write_index_tables`	Generate PyTables index datasets for all tables that contain an event_id or tel_id.
`write_muon_parameters`	Store muon parameters if available
`write_r0_waveforms`	Store R0 waveforms if available
`write_r1_waveforms`	Store R1 waveforms if available

Methods Summary

`__call__`(event)	Write a single event to the output file.
`finish`()	called after all events are done
`table_name`(tel_id)	construct dataset table names depending on chosen split method
`write_simulation_histograms`(event_source)	Write the distribution of thrown showers

Attributes Documentation

compression_level#: compression level, 0=None, 9=maximum

compression_type#: compressor algorithm to use.

contact_info#

A trait whose value must be an instance of a specified class.

The value can also be an instance of a subclass of the specified class.

Subclasses can declare default classes by overriding the klass attribute

context_metadata#: Additional metadata keywords and values that describe this data. This should be a dictionary where the keys will be appended to the CONTEXT section of the output file’s attributes. Keys can be hierarchical by using a space between each level, e.g. SIMULATION PRODUCTION would make a key PRODUCTION grouped under the key SIMULATION

datalevels#: returns a list of data levels requested

image_dtype#: A trait for unicode strings.

image_offset#: An int trait.

image_scale#: A float trait.

instrument_info#

A trait whose value must be an instance of a specified class.

The value can also be an instance of a subclass of the specified class.

Subclasses can declare default classes by overriding the klass attribute

output_path#: output filename

overwrite#: overwrite output file if it exists

peak_time_dtype#: A trait for unicode strings.

peak_time_offset#: An int trait.

peak_time_scale#: A float trait.

transform_image#: A boolean (True, False) trait.

transform_peak_time#: A boolean (True, False) trait.

transform_waveform#: A boolean (True, False) trait.

waveform_dtype#: A trait for unicode strings.

waveform_offset#: An int trait.

waveform_scale#: A float trait.

write_dl1_images#: Store DL1 Images if available

write_dl1_parameters#: Store DL1 image parameters if available

write_dl2#: Store DL2 stereo shower parameters if available

write_index_tables#: Generate PyTables index datasets for all tables that contain an event_id or tel_id. These speed up in-kernel pytables operations,but add some overhead to the file. They can also be generated and attached after the file is written

write_muon_parameters#: Store muon parameters if available

write_r0_waveforms#: Store R0 waveforms if available

write_r1_waveforms#: Store R1 waveforms if available

Methods Documentation

__call__(event: ArrayEventContainer)[source]#: Write a single event to the output file.

finish()[source]#: called after all events are done

table_name(tel_id)[source]#: construct dataset table names depending on chosen split method

write_simulation_histograms(event_source)[source]#

Write the distribution of thrown showers

Notes

this only runs if this is a simulation file. The current implementation is a bit of a hack and implies we should improve SimTelEventSource to read this info.
Currently the histograms are at the end of the simtel file, so if max_events is set to non-zero, the end of the file may not be read, and this no histograms will be found.