Patches and recent notes

In Syntax: Fix bug parsing nrm(ID).test and test.nrm(ID)In Core: Revised transformation of C matrix to Analysis order

The Windows and Linux download files are now much small because I have excluded the MKL libries that I was not using. Now all systems are less than 10 Mbyte. The Mac version is now as a .pkg file which should be easier to install. […]

Improved equation ordering; fix bug detecting groups. Allow random term ar1(row).ar1v(column) Fix overflow issue with large design matrix Fix Parsing bug with complex model term (us(Env.leg(X,3)) Fix bug setting error structures for multisection multivariate analysis

Unfortunately all previous Mac builds since have used an out of date routine to form the Ainverse. This issue was reported in build 109 and is fixed in build 110 available now. (Windows and Linux builds were not affected by this error.) Please advise any […]

Repaired pedigree reading bug introduced in build 106

Improved !TRIM filename field !SKIP k !KEEP g qualifiers to the PEDIGREE file line. Added !FOLDER path qualifier. Fixed !OUTFOLDER path qualifier. reduced confusion between DIRECTIVE (TAB/PLOT/VPR/…) names and short Y-variable names.

Added !TRIM filename field !SKIP k !KEEP g qualifiers to the PEDIGREE file line. Fix bug in operation of !Dv* transformation

Thanks for trialling Echidna. May God grant you blessing in 2020 through Jesus Christ This update includes* Implemented xfa1().geno.coru(T) variance structure * !WVR writes working variables to .ewv file * !SLOW qualifier reduces step size of variance parameter updates * Fixed bug in reported effects […]

Fix VPREDICT conversion of XFA to US !EQN 77 retrieves previous equation order if possible Fix bugs parsing Zero and xfa1(16) Revised parameter update process so iteration sequence has changed slightly.