Add cleanup script.

Long-needed code cleanup.
Better starting guess.

Add bzrignore.

Calculate S^{-1/2} via functional calculus, not via Cholesky.

Adapt for PyLinear op->comp/op split.

tag of inform@tiker.net--iam-2005/pywannier--iam--1.0--patch-16

Iterate longer, different bands-> no difference.

Updated to CG and tr[A*A^H]

Fixed-step *does* do some good

Integrated Marzari's gradient. Not what I had hoped for.
Sign seems wrong, too. Works otherwise.

Begun implementation of Marzari strategy.

wake of fempy changes

More revamp, fix visualization of Bloch functions.

Fixes to verify_* scripts.

Merge Fempy adaptation.

Various fixes before conversion to FEM k-space
Fixed compute_analytic_eigenmodes

Small improvements.
Removed arg() cruft.
Added initial Berghold mess.
Better initial guesses.

Berghold etc.
Sped up periodicization.
Began Berghold minimizer.

Boatloads of adaptations

camel case conversion fixes

Initial fixes due to PyLinear revamp

Update to match PyLinear changes

tag of ak@ixion.net--2003-diplomarbeit/pywannier--prototype--1.0--patch-50

final version

added verify_convergence.py

better (but not good) implementation of boundary interpolation

more visualization

visualizing 1D band diagrams works, except for a spurious factor of 0.5.

band diagram work

phase unification (no results)

fixed brown paper bag bug

draw exact meshes to gnuplot

fixed a cg bug, code cleanups

implemented polak-ribi`ere
  involved making k-dependent matrices an actual
    vector-space-like datatype (i.e. +,-,*)
  this functionality is abstracted out into a 
    tDictionaryOfArithmeticTypes in fempy.tools.
implemented separate view in band stack visualization
implemented interactive stopping

does it really work?
implemented scalar product fixes as suggested by new ideas

pre-doerfler on sep 2
u development over bz plot script
fix to scalar product

work at Josie's place
periodicity fixes, again
derivative debug facilities
more work on analytic modes

modularized eigenmode test, analytic eigenmodes

added interactivity levels

the "golden version" for doerfler