Release v1.2.2 of DynAdjust

Overview

This release contains several code enhancements and bug fixes, and addresses issues #15, #30, #78, #81, #80, #82, #85, #150 and #160. The following summarises the related (major) pull requests associated with these code changes.

General code enhancements

Numerous (too many to list) code enhancements have been applied to address issues identified by Codacy, Coveralls and Codecov. New (GitHub Actions) CI/CD scripts have been developed to invoke automated workflows on code push/pull request events:

• build a Ubuntu docker image
• build and test DynAdjust on Linux using cmake (debug and release variants).
• prepare for build on Windows using msbuild. This workflow is incomplete due to the complexities associated with building Intel MKL within a GitHub Actions workflow.

Installation scripts

• See #162 for a fix in FindXercesC.cmake, where installation failed to find a pre-existing installation of xerces-c in /opt/xerces-c/3.1.4.
• See #163 for code enhancements to make_dynadjust_gcc.sh, to streamline build behaviour across clone and no-clone options, and install_dynadjust_prerequisites.sh, which now cleans up any temporary folders created when downloading prerequisites.

Enhancements and bug fixes for import

• See #87 for changes to permit the sorting of measurements in *.m2s files.
• See #98 for enhancements to the application of a GNSS baseline scalars file.
• In relation to extending discontinuity renaming (#15):
  • See #88 for the initial changes to import to address this issue.
  • See #147 for final changes to import and adjust to complete the work.
  • Note that the changes applied in #147 change the structure of information written to the binary station file, meaning that *.bst files created by earlier versions of import cannot be reused by this or future versions of DynAdjust.
  • See #162 for a fix relating to the renaming of sites with discontinuities and every occurrence of a discontinuity site in the imported measurements (identified by @harry093 in discussion #149).
  • See #163 for further work to address a lingering renaming issue, to ensure that SNX sites are renamed using the start of the observation window defining the period up to which a discontinuity has been identified, rather than the date of the first SNX solution (which may vary depending on GNSS analyis).
• Address a DynaML station name parsing issue caused when two different station names (e.g. station-name-a and station-name-b) are provided within the one <DnaStation> record, as shown below. Whilst not strictly an import defect, parsing the second name (permissible by the schema) resulted in replacing the first name and thereby led to malformed station-to-measurement associations.

  <DnaStation>
    <Name>station-name-a</Name>
    <StationCoord>
      <Name>station-name-b</Name>
      ...
    <StationCoord>
  </DnaStation>

Enhancements and bug fixes for geoid

• See #89, #90, #94, #108, #109 and #146 for code changes, bug fixes and test data.

Enhancements to adjust

• See #83 for enhancements to address column overrun identified in issue #78.
• See #84 for a fix to incorrect iteration count identified in #82.
• See #86 for changes to prevent output of unused stations to the adjusted stations file, addressing #85.
• See #87 for changes to permit the sorting of the measurements to station table by measurement count.
• See #92 for changes to provide a capability to export adjusted station coordinates and uncertainties as a GNSS Y cluser.
• See #99 for an enhancement that prints memory consumption to the .adj and .dbg files on verbose execution.
• See #107 for a fix to a defect in the way adjusted measurements are printed on each iteration in staged mode
• See comments above in import (#147).
• See #162 for a fix addressing issue #160 and update corresponding formula in DynAdjust User's Guide.
• See #162 for a fix relating to archiving the estimated station uncertainties for simultaneous adjustments to *-rva.mtx. This defect led to strange values for SD(e), SD(n) and SD(up) in *simult.adj files when executing adjust in report results mode (i.e. --report-results).

Enhancements to plot and upgrade to GMT6

• See #137 for a fix to a defect in plotting single blocks.
• See #139 for code changes addressing #30.

DynAdjust User's Guide

• See #80, #138, #146, #160 and #162 for the updates applied to the User's Guide.

Acknowledgements

Special thanks to @harry093, @BatchelorJ, @nicgowans, @wapayze, @jameslevershadelwp, @stratiform and @tekwan for their contributions to tracking issues, reviewing algorithms and testing the release candidates.