Skip to content

Latest commit

 

History

History
84 lines (60 loc) · 4.72 KB

HISTORY.md

File metadata and controls

84 lines (60 loc) · 4.72 KB
Raw

Revision History for symlib

Rev. 2.0.4 (GLWH)

  • Changed pointer to allocatable in find_site_equivalencies so that UNCLE will compile with new symlib. This routine is only used in UNCLE (not used elsewhere in symlib or enumlib).

Rev. 2.0.3 (GLWH)

  • There were many small changes on my local copy that I committed, mostly changes in the absolute tolerances. They were made so long ago I don't remember why they were made.

Rev 2.0.2

  • Fixed a bug in both Smith Normal Form routines where it would attempt to divide by zero when one of diagonals was set to zero during one of the row/column operations.
  • Added an optional absolute tolerance to the minkowski reduce code.

Rev 2.0.1

  • At the last revision, there were some left over markers from merge conflicts. Removed those.
  • added 'regular = "true" to parameter tag for atom_pos; it had been changed from 'pointer' to 'allocatable'
  • removed 'atom_pos' from the unit test xml file
  • All of the unit tests seem to be passing fine except for two routines that are used in uncle. One is an out of bounds error in 'find_site_equivalencies'. The other is a missing file of some sort: ERROR: can't create representation for /Users/hart/codes/staging/symmetry.find_site_equivalencies.g/tests/find_site_equivalencies_nSites.2/nSites.out. File does not exist.. Will have to sort out those tests later.

2.0.0 Major revision (Jan 2019)

  • Changed some pointers to be allocatable arrays instead. This was to keep symlib compatible with enumlib which used allocatable arrays and general avoids pointers.
  • This counts as a major revision because it is a breaking change---the calling interface for several routines that are called from other programs are different now and the calling programs will have to change as a result.

Revision 1.1.4

  • Added a long int version of the SmithNormalForm algorithm.

Revision 1.1.3

  • Rewrote the SmithNormalForm algorithm to use a different pivoting strategy.
  • Added an optional output err_ flag to the SmithNormalForm alforithm to report an error but not stop if the code overflows.

Revision 1.1.2

  • Changed the bing_into_cell subroutine to use MOD instead of a do while loop so improve runtimes.

Revision 1.1.1

  • Fixed bug (reported in #19) where max and not maxval was being used when comparing arrays.

Revision 1.1.0

Revision 1.0.0

In rational_mathematics.f90, changed the comment for the SNF routine. The comments on the calling interface were incorrect (A and M had been backwards).

In symmetry.f90, added some "regular = true" flags in the xml comments.

Revision 0.0.1

Added unit tests for combinatorics.f90, it now has working unit tests for 66% of it's subroutines, and its supporting combinatorics.xml file. Also cleaned up the combinatorics.f90 file and adde XML documentation.

Added unit tests for compare_structures.f90, it has working unit tests for 100% of the subroutines used in the UNCLE code. There are 5 other subroutines that are not used that have not yet been unit tested. Also cleaned up and added XML documentation to the compare_structures.f90 file. Added the compare_structures.xml file as well.

Added unit tests for numerical_utilities.f90, it has working unit tests for 90% of its subroutines. Also changed the equal subroutines so that if the values are the same, it returns true no matter the tolerance that is used. Added the XML documentation to numerical_utilities.f90, the unit test files, and the numerical_utililties.xml file.

Added unit tests for rational_mathematics.f90, it has working unit tests for 60% of its subroutines. The last 37% are waiting on a fortpy bug to get sorted out.

Added unit tests for symmetry.f90, but it is still a bit messy. For the 10, subroutines only 2 have fully functional unit tests. The rest are rank3 arrays, or require logical inputs. There are a few, marked in the xml doc, that fail because the saved and new files have different precision apparently dp8 and dp4.

Added unit tests for utilities.f90, none work because of fortpy issues, but 100% of the subroutines have tests written.

Added unit tests for vector_utilities.f90. It is now 70% unit tested though the unit tests have some buggy floating-point issues.

Initial Repository: Revision 0.0.0

The first few commits were to get the repo up to scratch and nice and clean with installation instructions, etc. This includes the commit for new revision number 0.0.0. It includes an update of the *.xml files defining the unit tests so that they work with the distribution directory's structure.