Tags

Tags give the ability to mark specific points in history as being important

  • v0.2

    d3a49cbd · Improved installation · 
    WindowCovariance 0.2 — C++/SWIG implementation

Final state of the original C++ core before it was replaced by the pure-Python
Numba implementation in 7c656c6.

  C++ with Eigen and OpenMP, wrapped for Python via SWIG
  WindowCovariance/src/{Counts,Grid,Interpolator}.cpp
  WindowCovariance/include/*.h
  Required a compiler, SWIG, GSL and Eigen at install time

Superseded by v1.0.0, which computes the same quantities in pure Python with
Numba JIT kernels and needs no compiler. Archived on branch legacy/cpp-swig;
this tag marks the same commit as an immutable point of reference.

  • v1.0.0

    ea1d489b · Release v1.0.0 · 
    WindowCovariance 1.0.0

First public release.

Covariance of the galaxy two-point correlation function by Monte Carlo
integration over pairs, triples and quadruples of random-catalogue particles.
Pure Python with Numba JIT kernels; grid and octree backends; (r, mu) or
multipole output; HDF5 checkpointing for long MC runs.

BSD-3-Clause. Tested on Python 3.11-3.13 with numba 0.59-0.67 and numpy
1.26-2.5; 150 tests, including an independent brute-force check of C3/C4
against an exhaustive loop over raw coordinates.