# SrTiO3 (Pm-3m): the three VASP runs of `crystod --diagram --vasp`
#
#   crystod --diagram --co-left SrTi --co-right O3 --vasp .
#   crystod --diagram -c BAND/POSCAR   --co-left SrTi --co-right O3 --vasp .
#   crystod --diagram -c POSCAR-finish --co-left SrTi --co-right O3 --vasp .
#   crystod --diagram --co-left SrTi --co-right O3 \
#           --vasp ./BAND ./BAND_sublattice1 ./BAND_sublattice2
#
# Layout
#   BAND/                     the crystal run            Sr Ti O        NELECT 32
#   BAND_sublattice1/band/    the cation sublattice      Sr Ti Va2-     NELECT  8
#   BAND_sublattice2/         the anion sublattice       Va2+ Va4+ O    NELECT 24
#   POSCAR-finish             the SAME crystal on the OTHER origin (see below)
#
# The two settings
#   The three run POSCARs are written with Ti at the origin: Ti (0,0,0), Sr
#   (1/2,1/2,1/2), O (0,0,1/2) (1/2,0,0) (0,1/2,0), a = 3.90798883 A.  POSCAR-finish is
#   the relaxed cell as phonopy writes it, with Sr at the origin: Sr (0,0,0), Ti
#   (1/2,1/2,1/2), O (1/2,0,1/2) (1/2,1/2,0) (0,1/2,1/2), a = 3.90802055 A (8e-6
#   relative).  It is the same crystal with the origin moved by (1/2,1/2,1/2), and it is
#   here so that the test suite exercises the run-to-cell MAPPING of the engine: one
#   global origin shift, a permutation of the atoms (Sr Ti O O O -> ions 1 2 5 3 4) and
#   lattice-vector wraps, with the Va atoms mapping onto the removed ones.
#
#   Every number of the two analyses is the same -- raw and aligned energies, site
#   shifts, projected weights, purities, shares and connector weights agree to the last
#   printed digit.  What differs is the IRREP LABELS at R, X and M, which depend on the
#   choice of origin and are related one-to-one by
#
#     R : 1+ <-> 2- , 3+ <-> 3- , 4+ <-> 5- , 4- <-> 5+
#     X : 1+ <-> 3- , 2+ <-> 4- , 2- <-> 4+ , 5+ <-> 5-
#     M : 1+ <-> 4+ , 2+ <-> 3+ , 2- <-> 3- , 5+ , 5- fixed
#     GM: unchanged
#
#   The engine states which setting the labels belong to in the terminal report, in the
#   level table, in the JSON ("setting") and as a chip on the page.
#
# What was trimmed
#   Real output of three finished VASP 6.5.1 runs (PAW/PBE, ENCUT 550 eV, LORBIT 12, a
#   6x6x6 irreducible mesh plus the zero-weight special points).  PROCAR and EIGENVAL
#   were restricted to the four k points the diagram reads -- GM, X, M and R -- and
#   their headers rewritten to match; the complex LORBIT = 12 phase tables are
#   untouched, so the projections, irreps, shares and connector weights are the real
#   ones.  The files are gzipped (every reader of crystod/vasp_io.py accepts a .gz
#   sibling), and the whole fixture is about 82 kB instead of 45 MB.
#
#   The OUTCARs are excerpts: the run parameters, the Fermi level, the CrystOD Va
#   point-charge table, and the two lines that name each PAW dataset and its valence
#   configuration.  The PAW header tables of a real OUTCAR are a verbatim dump of the
#   POTCAR header and are NOT reproduced here; no POTCAR file is shipped, and nothing
#   from the VASP source tree is.
#
# Fidelity
#   Every number the test suite checks against this fixture is the number the untrimmed
#   288-k-point runs give.
#
# The full runs, the modified VASP that understands the Va species, and the patch
# itself are not part of CrystOD; VASP is proprietary and the patch is available to
# VASP licensees on request.
