Chemical routines based on libtorch

We develope this library based on the LTS pytorch build 1.8.2, so using Torch-Chemistry along with libtorch 1.8.2 is recommended

Gaussian:

• value of gaussian functions
• product of gaussian functions
• gaussian integrals
• multivariate gaussian random number

Polynomial:

• value of polynomials
• Jacobian of polynomials
• transformation of polynomials under coordinate rotation
• transformation of polynomials under coordinate trasnlation
• symmetry adaptation

Chemistry:

• alternative representation near electronic degeneracy
• phase fixing for quantum observable matrices
• normal mode analysis

Internal coordinate:

• Cartesian -> internal coordinate
• Jacobian of internal coordinate over Cartesian coordinate
• 2nd order Jacobian of internal coordinate over Cartesian coordinate
• Cartesian coordinate gradient <-> internal coordinate gradient
• Cartesian coordinate Hessian <-> internal coordinate Hessian
• for more details see intcoord.md
• for symmetry adaptation and scale see SASintcoord.md

Linear algebra:

• outer product for general tensors
• map a vector to a symmetric tensor
• matrix dot multiplication between 3rd-order tensors
• matrix dot multiplication between 4-th and 3rd-order tensors
• matrix outer multiplication for general tensors
• unitary transformation

FORTRAN:

• generalized eigenvalue problem solver (dsygv)

Utility:

• some general basic routines

1. mkdir build lib
2. cd build
3. cmake ..
4. cmake --build .
5. mv lib* ../lib
6. add include and lib to your path

#include <tchem/tchem.hpp>

1. libtorch
2. Cpp-Library