Skip to content

Latest commit

 

History

History
131 lines (98 loc) · 4.56 KB

RELEASE.rst

File metadata and controls

131 lines (98 loc) · 4.56 KB
Raw

0.21 - 2012-12-09

  • Add three new content types: SIM SEA, EPHIN housekeeping, and CPE.
  • Numerous infrastructure improvements in ingest and update process.

0.20 - 2012-08-11

  • Use weighted mean and stddev for calculating stats.
  • Use float64 to accumulate sum for computing stats mean.
  • Rebuild stats files for the full mission.
  • Fix bug that number of samples for daily stats was incorrect.
  • Add notes and regression testing code for re-building stats files.
  • This release fixes issues #38, #39, and #41.

0.19.1 2012-06-22

  • Add MSID.interpolate() method which is like MSIDset.interpolate()
  • Speed up interpolate() methods using the new Ska.Numpy.interpolate.
  • Add MSIDset.filter_bad_times() method that applies the bad times filter to all MSIDs in a set.
  • Speed up filter_bad_times() by using a single mask array over all bad time filters.
  • Add some unit / regression tests.

0.19 - 2012-05-04

  • Fix MSID.raw_vals() so it handles state codes with different lengths
  • Fix problem in iplot where units not tied to fetched MSID
  • Add units to DEA housekeeping MSIDs

0.18 - 2012-04-30

  • Make it possible to reliably use the import mechanism to select different unit systems within the same script or Python process with no interactions.

0.17 - 2012-04-17

  • Improve tutorial documentation
  • Modify PCAD derived parameters to use only predictive ephemeris
  • Redefine DP_ROLL_FSS and DP_PITCH_FSS to improve accuracy
  • Allow for filter_bad_times to function for data that have already been filtered.

0.16 and 0.15

A number of major new features are available in the release 0.16 of the Ska engineering archive:

  • Built-in plotting capability and an interactive plot browser that allows arbitrary zooming and panning. See tutorial sections: Plotting time data (http://goo.gl/6tGNZ) and Interactive plotting (http://goo.gl/2dG4c)
  • Support for plotting state-valued MSIDs and for accessing the raw count values. See: State valued MSIDs (http://goo.gl/9R0Tz)
  • Support for accessing Telemetry Database attributes related to an MSID. See: Telemetry database (http://goo.gl/pPo0s)
  • PCAD derived parameters (main code from A. Arvai). See: Derived PCAD parameters (http://goo.gl/iKDUK)

API Changes

  • MSIDset.interpolate() now behaves more intuitively. This was done by making the times attribute of interpolated MSIDs correspond to the new (linearly-spaced) interpolated times. Previously times was set to the nearest-neighbor times, which is not generally useful. See #20 and Interpolation (http://goo.gl/5U5Kp)

Bug fixes

  • Fixed problem where Msidset fails for 5min and daily values.

0.14

Minor updates and bug fixes:

  • Change max_gap for ACISDEAHK from 3600 to 10000 sec
  • Add midvals attr for stat fetch
  • Fix ss_vector() to use quaternion midvals and handle missing telemetry
  • Fix typo in fetch.logical_intervals
  • Explicitly set --data-root when calling update_archive.py in task_schedule.cfg

0.13

Version 0.13 of the Ska.engineering archive contains a number of new features:

  • Support for derived parameter pseudo-MSIDs. Currently there are number of thermal values and ACIS power values.

  • Two new fetch.MSID methods:

    • logical_intervals() returns contiguous intervals during which a logical expression is true.
    • state_intervals() determines contiguous intervals during which the MSID value is unchanged.

  • Two new classes fetch.Msid and fetch.Msidset. These are just like fetch.MSID or fetch.MSIDset except that filter_bad=True by default. You can use the new classes just like before but you'll always only get good data values.

  • Changed definition of the "vals" attribute for '5min' and 'daily' stat values so "vals" is now the sames as "means". Previously the "vals" attribute for a statistics fetch was the exact telemetry value at the interval midpoint. This quantity is not that useful and prone to errors since one frequently does things like:

    dat = fetch.MSID('tephin', '2011:150', stat='5min')
plot_cxctime(dat.times, dat.vals)

  • Caching of previously fetched data values. This is disabled by default but used for telemetry and derived parameter ingest. In certain circumstances caching can be useful.

  • New function for truncating the engineering archive at a certain date. This is useful for fixing the database in the event of a corruption.