HiPERCAM pipeline news

This page documents important changes in the pipeline, as a help to whether you should update.

  1. 05 May 2021, Version 0.22.0: matplotlib-based real time image plotter introduced

    • rtplot, matplotlib-based version of rtplot. Allows panning and zooming during the plots, choice of colour map, and profile fits from multiple CCDs.

    • setaper, hplot, setdefect also now support colour maps

    • averun output name now synchronised with input

    Here is the full list of changes, 0.21.2 to 0.22.0.

  2. 17 March 2021, Version 0.21.2: bug fix for a problem with saving parameter default files.

    • new environment variable HIPERCAM_DEFAULT_SOURCE introduced

    • makebias output name synchronised with input

    • lots of minor improvments to logging scripts

    • filtid, new command, aimed at potential filter ID from sky flats

    Here is the full list of changes, 0.21.1 to 0.21.2.

  3. 05 March 2021, Version 0.21.1: again a large number of updates and fixes in this version. Most are small or to do with data logging, but the cumulative effect is significant and some bugs have been fixed. Tseries in particular has undergone a bit of an overhaul and has expanded functionality with an improved handling of bitmasks. Some of these may break existing code based upon Tseries; apologies. Main changes:

    • makedark bug fixed (was not subtracting bias)

    • new routine ‘flagcloud’ added to flag reduce logs interactively

    • reduce argument order and handling improved

    • a couple of new formats added to fits2hcm

    • Tseries: much better handling of bad and flagged data

    • Tseries bin option debugged

    • plog now has an option of specifying a plot title

    • should now be possible to display images with inverted axes

    • fixed bug preventing ‘splice’ from working

    • fixed bug for handling ULTRASPEC drift mode

    Here is the full list of changes, 0.20.0 to 0.21.1.

  4. 10 July 2020, Version 0.20.0: a large number of updates and fixes in this version. Apart from fixes to problems that would have been obvious as they were show-stoppers, the main changes are:

    • greatly improved target location for small windows

    • fixed bug with target location and parallel processing

    • added option to limit maximum beta for Moffat fits

    • added various specialist routines to do with timing

    • started on a routine ftargets which uses ‘sep’ to automate reduction

    There are many changes here, and I expect there to be bugs as a result, but some of the changes are significant enough that it is better to release now rather than wait longer. See also the full list of changes, 0.19.8 to 0.20.0

  5. 26 February 2020, Version 0.19.8: there have been many changes since the 0.16 versions. Apart from many bugfixes, the main ones are:

    • dark subtraction implemented

    • significant speed-ups implemented,

    • trimming of row & columns for ULTRACAM,

    • better display level setting through region definition,

    • ability to ignore frames with bad times in reduce,

    • rtplot bias level warnings added,

    • psf_reduce added for crowded fields.

    • rupdate script added to update old reduce files

  6. 26 June 2018, Version 0.16.XX: there have been very many recent changes in the pipeline as I have started to use it in earnest for reduction and have fixed a few problems encountered along the way. I would strongly advise anyone using the pipeline to update to the latest version. Above all I have improved the robustness of the aperture positioning making it less vulnerable to cosmic rays, removing an irritating overflow problem associated with Moffat profiles, and putting in several more checks to guard against positioning problems. Altogether these significantly reduce the chances that you lose apertures during a reduction.

    I have also implemented an experimental parameter fit_alpha which applies a fraction of the shift determined for an aperture relative to any references which is equivalent to smoothing the position differences.