Release Notes¶
main¶
Feb. 20, 2024 don la dieu (nega at icecube.umd.edu)¶
IceTray Release v1.9.2
Delete mailinglist files (#3357)
Fix extension for various files (#3330)
Add upgrade rules to ruff (#3316)
Dec. 14, 2023 don la dieu (nega at icecube.umd.edu)¶
IceTray Release v1.9.0
general python cleanups (ruff/E712) (#3268)
fix E703 useless-semicolon (#3266)
June 22, 2023 don la dieu (nega at icecube.umd.edu)¶
IceTray Release v1.7.0
Make I3Tray.I3Tray and icecube.icetray.I3Tray the same thing. (#3102)
Dec. 12, 2022 don la dieu (nega AT icecube.umd.edu)¶
IceTray Release v1.5.0
Documentation cleanups
Jul. 27, 2022 don la dieu (nega AT icecube.umd.edu)¶
IceTray Release v1.4.0
Drop support for Python 2
Updates for C++17
Clean up documentation
Feb. 22, 2022 don la dieu (nega AT icecube.umd.edu)¶
IceTray Release v1.3.0
Merge the two different finiteReco converters
Documentation updates
Dec. 20, 2019 Alex Olivas (aolivas@umd.edu)¶
Combo Release V00-00-00
November 29, 2017 Kevin Meagher (kmeagher@ulb.ac.be)¶
Release V01-05-05
Fix tests: use GCD file from $I3_TESTDATA
doc cleanup: remove dox file
script cleanup: remove TrashCan and Finish
March 24, 2017 Kevin Meagher (kmeagher@ulb.ac.be)¶
Release V01-05-04
Add serialization dependendency
Cleanup after unit tests
Add OutputName to gulliver modules
May 2, 2016¶
Release V01-05-03
moving I3LineFitParams from finiteReco to recclasses
– Hans Dembinski
updated documentation: ported wiki pages from Aachen to wiki.icecube.wisc.edu, created rst project documentation, reviewed inline doxygen comments
– Sebastian Euler
Removed LinkDef.h which really shouldn’t be needed anymore. If something ROOT-y is broken, this could be why.
– Alex Olivas (?)
April 3, 2015 Meike de With (meike.de.with@desy.de)¶
Release V01-05-02
Delete project-config.mk which is outdated
– Don la Dieu
Small code changes to solve compiler warnings
– Sebastian Euler
Recognize I3_TESTDATA
– Alex Olivas
December 9, 2014 Meike de With (meike.de.with@desy.de)¶
Release V01-05-01
Add name to the Delete/Rename modules in the tray segment
– Marcel Zoll
September 29, 2014 Meike de With (meike.de.with@desy.de)¶
Release V01-05-00
Remove templating that was necessary for being able to use RecoHits or RecoPulses. Now that finiteReco does not accept RecoHits anymore, also the templating is not needed anymore. Adapt the test accordingly.
Remove obsolete functionality for using RecoHits instead of RecoPulses
Correct a bug in I3FiniteCalc.cxx (references to temp objects) which caused tests on Ubuntu 14.04 to fail
Cleaning up code and adding documentation
– Sebastian Euler
July 11, 2014 Meike de With (meike.de.with@desy.de)¶
Release V01-04-02
Change criteria to avoid IceTop DOMs in currently considered extensions which might increase the number of DOMs per string
– Joao Pedro Athayde Marcondes de Andre
Change tray segments to allow multiple instances of the same segment within one script
– Sebastian Euler
April 17, 2014 Meike de With (meike.de.with@desy.de)¶
Release V01-04-01
Port the ‘If’ option to the tray segment
– Marcel Zoll
Change vertex shift time in the advancedLengthReco segment from TFirst to TNone
– Sebastian Euler
January 13, 2014 Meike de With (meike.de.with@desy.de)¶
Release V01-04-00
Fixed a bug that would result in DetectorLength > 1.5 km, caused by IceTop DOMs not being filtered out. Now, DetectorLength is calculated using only DOMs with omtype I3OMGeo::IceCube.
– Marius Wallraff
Added tray segment for the length reconstruction
– Sebastian Euler
August 21, 2013 Meike de With (meike.de.with@desy.de)¶
Release V01-03-04
Import dataio
– Jakob van Santen
Explicitly use boost::shared_ptr for libc++
– Claudio Kopper
Replace string exceptions with real exceptions
– Nathan Whitehorn
Clean up deprecation warnings related to new dataclasses release
– Meike de With
February 21, 2013 Emanuel Jacobi (emanuel.jacobi@desy.de)¶
Release V01-03-03
Removed dependency on log4cplus
– Nathan Whitehorn
Added a check for empty pulse series to I3FiniteCalc -> log_warn
– Marius Wallraff
Removed photonics dependancy
Removed hard coded constants
Cleaned up warnings generated by clang
– Kevin Meagher
Removed dependency on DOMcalibrator and FeatureExtractor
–Emanuel Jacobi
Reduced loglevel on request of PnF
– David Boersma
September 11, 2012 Emanuel Jacobi (emanuel.jacobi@desy.de)¶
Release V01-03-02
Use GetName() instead of getting a configuration from the context
– Nathan Whitehorn
Require dependency on python (I3.h depends on Python.h)
– Fabian Kislat
June 25, 2012 Emanuel Jacobi (emanuel.jacobi@desy.de)¶
Release V01-03-01
Fixed a bug in the calculation of the hit probability inside PhPnhPhotorec.cxx, that completely degraded the length reconstruction and caused the track length to be massively overestimated. (Not entirely sure how, though…)
– Sebastian Euler and Marius Wallraff
Fixed a bug in the photonics call inside PhPnhPhotorec.cxx: The particle speed was passed in m/ns, but PhotonicsSource expects a value in fractions of c.
The physics impact of this bug seems to be confined to monopole reconstructions, as the speed parameter in PhotonicsService is ignored for muons and cascades.
– Sebastian Euler and Marius Wallraff
Prevent compiler warnings
– Claudio Kopper
January 11, 2012 Emanuel Jacobi (emanuel.jacobi@desy.de)¶
Release V01-03-00
In PhPnhP*.cxx, the manual calculation of the Poisson distribution was replaced by a boost implementation, making unecessary the use of long doubles which caused the compilation on FreeBSD 9 systems to fail.
Two minor physics implications:
The new implementation still yields correct results when the old one failed due to too large arguements in E^NHit (leading to a hit probability of zero in this DOM). This should have happened rarely at most because it would have caused warnings.
This fixed a bug in PhPnhPhotorec.cxx that was introduced in June 2009, where gamma(n+1) was replaced by (n+1)! instead of n!.
A recently depricated PhotonicsService::SelectSource method has been replaced by non-deprecated candidate function using a PhotonicsSource object. The new version takes into account the speed of the input I3Particle, which can lead to different results for tracks where v!=c.
– Sebastian Euler and Marius Wallraff
Use new SetGeometry Method
– Robert Franke
Fixed bug in I3FiniteCutsConverter
– Markus Voge
December 15, 2011 Emanuel Jacobi (emanuel.jacobi@desy.de)¶
Release V01-02-00
Fixing the bug in I3GulliverFinitePhPnh introduced with r70265: While before that revision event durations were overestimated by a factor of 10, our fix caused the default event duration of 10 µs being used for every event (due to a misconception of what I3Time::GetModJulianNanoSec() is doing). Thus the event durations were now systematically underestimated.
Previously, the hit probability could become 1 if the track came very close to a DOM. This resulted in a LLH value of -inf. If the hit probability is 1, now 1-numeric_limits<double>::epsilon() is returned instead.
– Sebastian Euler
Added ENSURE_EQUALS
– Marius Wallraff
Fixed unit test
– Nathan Whitehorn
August 9, 2011 Emanuel Jacobi (emanuel.jacobi@desy.de)¶
Release V01-01-00
Bugfix:
Updated tests
Adapted to various changes in offline-software
– several authors
March 12, 2011, Fabian Kislat (fabian.kislat@desy.de)¶
Release V01-00-00
Fixed a bug in I3GulliverFinitePhPnh: Previously, the service used DAQ times and interpreted them as nanoseconds, overestimating durations by a factor of 10. This fix effectively reduces every DOM’s noise rate by a factor of 10.
Additionally, the default noise rate was reduced from 1000 Hz to more realistic 500 Hz, and the default event duration (only used in emergencies) was increased from 5 µs to 10 µs.
– Sebastian Euler et al.
Use I3Reader instead of I3ReaderService in tests.
– Jakob van Santen
November 06, 2010, Fabian Kislat (fabian.kislat@desy.de)¶
Release V00-03-01
Fixed situations where I3StartStopLProp would miss to write its output frame objects.
Fixed PhPnhMCDist unit test.
Added file operation checks to MCDist, preventing seg faults and compiler warnings.
Switched more log_warns to log_infos, added a counter for events for which I3StartStopPoint failed due to bad reconstructions that have FitStatus OK (which happens quite frequently since OK only means that the fit converged to a local likelihood maximum). If this happened at least once, a summary log_warn is given at I3StartStopPoint’s (new) Finish method.
– Sebastian Euler et al.
October 29, 2010, Fabian Kislat (fabian.kislat@desy.de)¶
Release V00-03-00
Switched I3Modules to I3ConditionalModules.
– Andreas Gross
Fixed the pybindings.
Removed ROOT dependency (TMath::Factorial).
Introduced various checks for NULL pointers.
Introduced fit status checks, suppressing log_warns in favor of log_infos for failed fits.
– Sebastian Euler et al.
August 17, 2010, Fabian Kislat (fabian.kislat@desy.de)¶
Release V00-02-01
Remove PSInterface dependency from test scripts because they do not need it
– Fabian Kislat
March 11, 2010, Fabian Kislat (fabian.kislat@desy.de)¶
Release V00-02-00
Bugfix:
Corrected bug that led to identical likelihood values
– S Euler
Feature:
added parameter DefaultEventDuration
– S Euler
November 08, 2009, Fabian Kislat (fabian.kislat@desy.de)¶
Release V00-01-01
Bugfix:
One test failed because of a typo
– Fabian Kislat
October 26, 2009, Fabian Kislat (fabian.kislat@desy.de)¶
Release V00-01-00 (initial release)
Features:
it’s reviewed
– Jan-Patrick Huelss
Improvements:
Added python bindings
– Jan-Patrick Huelss
