(At the moment a bit slow, optimization with Cython seems necessary).

Replaced res with a for the grid spacing
(resolution is a reserved word in electron microscopy
Work in progress: jacobi matrix for projection

phasemapper: added Kernel helper class with methods to calculate the jacobi
matrix or to multiply a vector with the (transposed) jacobi matrix without
explicitly calculating the full matrix
magcreator: added Shape.ellipsoid
numcore: added (experimental) speed-up for jacobi-matrix (not faster)
scripts: modified get_jacobi to test the new Kernel class,
tests: test_compliance now also searches and creates a list of TODOs

phasemapper: new function get_kernel() for convenience, mag_real_fast() for
calculation of the phase with pre-computed lookup-tables
projector: single_tilt_projection() now works as intended
Added scripts for the creation of core-shell-structures
Added script for charge calculation by contour integration
Added script for creating tilt series of core-shell-structure for reconstruction
Various minor modifications, mostly nomenclature and streamlining

pyramid: all docstrings in numpydoc (package installed) format (used by spinx)
holoimage: removed error, phase was falsely multiplied by pi/2
magcreator: removed create_mag_dist_comb, renamed functions for disc and vortex
magdata: Allows creation of empty MagData objects and easy adding of magnitudes
phasemap: Allows units to be set for plots, added 3D plot
phasemapper: Electrostatic component implemented
projector: Projection now includes thickness profile
docs: support for html, latex and pdf (experimental), use "make *" in shell
scripts: new structure, deleted some compare_* functions, added test_methods
.hpignore: now also ignores *.c and _* files (_build used in docs)

holoimage: changed colorwheel (now without axes)
magcreator: create_vortex now also allows a 3D-center (2D is extracted)
magdata: new methods: get_mask() and scale_down()
phasemap: changed plotting defaults (new kwargs)
phasemapper: started on electrical contribution

setup.py: Final cleanup
deleted initpyximport, compare_method_errors and get_revision (now in setup.py)

holoimage: turned off interpolation
test_compliance: added scripts to be tested
setup: work in progress

numcore: deleted c1, c2, c3, name changes in phase_mag_real
analytic: name change beta to phi
phasemapper: changes to include numcore
test_compliance: also checkes .pyx files now
setup: now works properly (hopefully on other systems, too)
compare_discs: added script to compare discs

holoimage: densitiy default set to 1

magcreator: Added the second angle for the magnetization
reconstructor: Changed method name
all others: Changed "beta" to "phi" to reflect the new angle in magcreator

test_compliance: Added PEP8 TestCase

scripts: changed scripts to work with the new syntax of the package
PhaseMag and MagData: added load and save options for NetCDF files
MagData: changed 3D-plotting to use mayavi (faster and better looking)
test: added more TestCases for various modules
phasemapper: added computation method to use mx and my (now standard)
compare_method_errors: first draft of script to compare the different errors

Scripts are nearly adapted to the new structure and easier to read.
(Except analytic module and scripts get_jacobi and inverse)

For Pushing Purposes

Changed setup.py
Implemented assertion to check if jacobi matrix is the same

test: changed the dataloader TestCase
pyramex: first experiments with Cython and SCons
dataloader: changed the order in which the dimensions are saved,
        z_len is no longer part of the magnetization (this was the reason why
        we had to divide by res, which is equal to z_len for calculating coeff)
magcreator: shapes are now separate functions, mag_shape is now an argument for
        one unified create_hom_mag-function
phasemap: unification of the phasemap-functions in real space (just the pixel-
        field is calculated different now)

Removed old scripts (AnalyticSolution and DataLoader)
Added several template testcases

jacobi matrix now contains the derivatives instead of descriptive strings