Version 1.2.8 (fixes, ehnacements, improved datum management)

[rogerfraser]

Overview

This pull request addresses several defects and shortcomings, improves the process for handling reference frame and epoch throghout, and offers various code improvements across all applications and documentation, including:

• Fix an issue with sorting adjusted GNSS measurements according to the N-statistic. This resolves N-Stat sorting improvement for Type X measures (and others?) #207.
• Fix an issue with reporting the results of a previous adjustment. This fixes Zero variance matrix components reported by adjust.exe when --report-results is run on phased-staged networks #221.
• Several changes have been introduced to streamline the way reference frame and epoch are handled throughout all applications. This closes Reftran --reference-frame argument to be optional (adopt existing project datum and epoch) #223. See note below.
• Fix the exporting of ignored measurements. This fixes Import --export-dna-files from DynaML to DNA format does not export ignored directions to DNA format #225.
• Fix a problem with reporting measurement ID and cluster ID when reporting adjusted GNSS measurements local reference frame. This fixes Measurement ID and Cluster ID are erroneously reported when "--output-adj-gnss-units 1" is provided to adjust.exe #227.
• Make the language consistent across DynAdjust applications and the User Guide for Zenith Distance. This addresses Review documentation and code outputs for consistent application of 'Zenith Distance' vs 'Zenith Angle' #229.
• Eliminate data type ambiguities by removing namespace includes for std and boost, ensuring full support for compiling with C++20. This addresses C++ 17 and C++ 20 build issues due to ambiguity of data type names #230.
• Fix a dnaplot defect using adjusted positional uncertainty files. This fixes dnaplot error #231.
• Revisions to the User's Guide to account for changes introduced by issues 223 and 229

Important notes on changes to default behaviour

Users should be aware of two major changes to the way DynAdjust handles the application of datum (reference frame and epoch):

• import allows the project reference frame to be set by an optional argument (--reference-frame) OR the reference frame label in the header of the first input file.
• import now has an option to provide a project epoch (--epoch).
• • New warning messages have been introduced to import to alert the user when the input data is aligned to multiple (inconsistent) reference frames.
• reftran no longer requires, as a mandatory option, reference frame (--reference-frame) to which all stations and measurements are to be aligned. In this case (and unless provided upon executing reftran), the project reference frame and epoch set by import will be used.

Acknowledgements

Special thanks to @nicgowans, @jhaasdyk-au, @jasonwheritage, @neilebrown, @alas2 for their help in identifying issues, offering suggestions, and/or testing code changes.