Skip to content

Latest commit

 

History

History
82 lines (65 loc) · 2.59 KB

CHANGELOG.md

File metadata and controls

82 lines (65 loc) · 2.59 KB

v2.3

  • Support for anisotropically-weighted fields (#217)

v2.2.3

  • Bugfix: saving norm_type in workspace fits files (#212)

v2.2.2

  • Improved bug fix (#204)

v2.2.1

  • Fixed small bug due to floating point error in CAR initialisation (#204)

v2.2

  • More convenient constructors for workspaces and FKP normalisation (#201)

v2.1

  • Catalog-based pseudo-C_ells (#184)
  • Tutorial notebooks and improved documentation (#192, #195, #196, #197)
  • Unit tests for mac-OS (#194)

v2.0

  • Improved error messages (#187)
  • Migrated to ducc from libsharp (#183)

v1.6

  • Bugfix: sign of Wigner 3j symbols (#180)
  • Faster method for spin-0 Wigner 3j symbols (#178)
  • No deprojection bias for lite fields (#177)
  • Bugfix for rectangular covariances (#175)

v1.5.1

  • Bugfix in Toeplitz calculation (#169)
  • Bugfix for rectangular covariances (#175)

v1.5

  • Compilation refactor (#158, #162)
  • Updated installation instructions for NERSC (#160)
  • Ensure correct mask endianness (#161)

v1.4

  • Option to read only the binned MCM (#151)

v1.3

  • Spin-0-only covariance workspaces (#147)
  • Added functionality to return field mask (#142)
  • Fixed a couple of bugs (#147, #144, #140)

v1.2

  • Uses most recent version of libsharp (libsharp2) (#121)
  • Various bug fixes (#123, #124, #126)
  • Added functionality to return the field's alms (#131)
  • Migrated tests to pytest and github actions (#132)

v1.1

  • Faster MCM via symmetry (thanks to Thibaut Louis) (#111)
  • Covariance calculable for coupled spectra (#113)
  • Arbitrary spins (#115)
  • Additional binning functionality (thanks to Joe Zuntz) (#116)
  • Lightweight and map-less fields (#117)

v1.0.2 changes

  • Pip-installable (#95).
  • Installation instructions on rtd (#95).

v1.0.1 changes

  • Input maps and templates can be masked (#93).
  • Input theory power spectra are culled to lmax automatically (#90).
  • Libsharp, HEALPix and libnmt automatically installed (#84).
  • Code is now PEP8 compliant (#86).
  • Workspaces are now stored as FITS files (which should make them easier to read externally) (#83).
  • More convenient NmtBin constructors (#80).
  • Lmax_mask, lmax_sht, lmax_bins (#74).
  • Parallelized Gaussian covariance (#73).
  • Ability to recover bandpower window functions easily (#55)

v1.0 changes

  • Support for rectangular pixelizations (CAR pixelization with Clenshaw-Curtis separation).
  • Support for spin-0 and spin-2 covariance matrices.
  • Convenience functions to recover bandpower windows and the unbinned mode-coupling matrix.
  • Arbitrary ell-weighting (e.g. can compute D_l = l*(l+1)*C_l/2pi).
  • Simultaneous calculation of the MCM for all spin-0 and spin-2 combinations.