Releases: psi4/gau2grid
Releases · psi4/gau2grid
Release list
v1.3.0
v1.3.0 release. This release focuses on compatibility across different architectures and compilers in addition to new computers for orbitals-on-a-grid which is up to 10x faster than computation of a collocation matrix following by a summation.
Notes:
- Adds direct orbitals-on-a-grid computation.
- New documentation website: https://gau2grid.readthedocs.io/en/latest/
mm_mallochas been changed to the pragmaALIGNED_MALLOCfor greater compiler and platform compatibility.- The precomputed RSH coefficients now use the Python Decimal module for greater platform compatibility.
- Added PGI compilers to the compiler comparability suite.
- Inner loop optimizations from VTune recommendations.
- Adds Molden cartesian output ordering.
v1.2.0
v1.1.0
Full v1.1.0 release.
Notes:
- Gau2Grid can now be compiled with different cartesian and spherical output orders:
-DCARTESIAN_ORDER- The cartesian ordering of the basis functions (row)
-DSPHERICAL_ORDER- The spherical ordering of the basis functions (cca, gaussian)
v1.0.1
Full v1.0.1 release.
Notes:
- setup.py Can now compile and install a full C version of gau2grid.
- Added versioneer.
- Compiler splits higher angular momentum into smaller loops for more efficient vectorization and reduction of compile time.
- Auto decision of cache alignment sizes.
- Saves RSH coefficients to file so mpmath is no longer required.
- Expands testing matrix to more LLVM and GCC versions.
Patch:
- CMake version is now correct.
v1.0 Release
Full v1.0 release.
Notes:
setup.pyCan now compile and install a full C version of gau2grid.- Added versioneer.
- Compiler splits higher angular momentum into smaller loops for more efficient vectorization and reduction of compile time.
- Auto decision of cache alignment sizes.
- Saves RSH coefficients to file so
mpmathis no longer required. - Expands testing matrix to more LLVM and GCC versions.
v0.1 Beta Release
Advertised Version: 0.1
Continuous Version: 0.1
Release Date: 13 Dec 2017
Documentation: N/A
Availability: Public, GitHub source, CMake build
Features:
- Arbitrary angular momentum gaussian to cartesian collocation grids.
- Spherical transforms available.
- NumPy and optimized C references available.
- Naive and fast rectangular transpose functions.
- Fully test suite.