Skip to content

Latest commit

 

History

History
56 lines (49 loc) · 2.35 KB

UPDATES.rst

File metadata and controls

56 lines (49 loc) · 2.35 KB
Raw

Updates

0.3.3

  • The underlying datamodelbase module is now moved to the new separate yabadaba package. Tons of minor updates associated with this and new yabadaba features.
  • Random bug fixes.

0.3.2

  • bad_lammps_potentials list added which lists the LAMMPS potentials entries that should always issue errors if used. This is useful for anyone wishing to test both old and new versions of potentials by helping to process calculation errors.
  • More updates for KIM potentials - now any KIM models associated with multiple potential entries will generate separate PotentialLAMMPS records for each. This simplifies the handling of these edge cases and helps avoid compositions associated with untrained cross-interactions.
  • Citation records now treat "journal" as optional when loading. CrossRef bibtex seems to now be missing this value?

0.3.1

  • Query methods are no longer static methods to make them consistent with other operations.
  • Retrieve methods added that allow for a record to automatically be saved as a file after it was queried.
  • The list of most stable isotopes added to atomic info to avoid having to specify isotope numbers for radioactive elements.
  • Improvements to cache handling for local databases.
  • The order of parameters for database methods were rearranged for consistency and practicality.
  • All default query parameter values are now None to avoid unwanted automatic filtering of records.
  • Additional bug fixes identified during testing

0.3.0

  • Major overhaul of the package. Most of the code was reworked to add functionality and streamline database interactions between the potentials, atomman and iprPy packages.
  • Basic database and record handling now in a separate sub-package, which is planned on being fully branched off later.
  • A csv-based caching system is added for local style databases to speed up queries.
  • Records now specify recognized query parameters allowing for custom "fast" queries to be generated for the different database styles.
  • Loading of record styles no longer necessary.
  • Performance improvements related to how the query actions are performed.
  • The handling of openKIM models has been improved.
  • Tools relating to building and analyzing EAM parameter files have been added in.
  • Specific database operations now exist for all supported record styles.