Previously it was never explicitly stated that the wavefunctions
returned by modes() are only defined over the unit cell of the
infinite system. This was confusing, as syst.sites contains
not only the unit cell sites, but also the connected sites in
neighboring cells.

Closes #178.

As of version 1.1.1 Sympy does not include MutableDenseMatrix
in 'all_classes', so expressions containing such terms did not trigger
the warning in kwant.continuum.sympify when 'locals' are provided.

Closes #180.

Previously, when plotting a 3D system with 'site_size' provided
explicity as a sequence, a ValueError would be raised when using
'%matplotlib inline' or '%matplotlib notebook' from a Jupyter
notebook. We do not add a test for this, because that would require
depending on Jupyter.

Closes #39.

This backend does not require a display to be connected,
which is good for tests.

Closes #83.

The lattice constant is never changed from 1, and in addition can be
varied independently of the 't' and 'α' parameters; This should not
be the case if the model is meant to be derived from a discretization
of a continuum model. We remove the lattice constant to limit confusion.

For a spin-orbit term of the form '-iασ_y∂_x' the hopping
for 'HoppingKind((1, 0), lat)' (*from* 'x' *to* 'x + a')
is 'iασ_y/2'.

Closes #175.

Previously, an erroneous current was shown at *zero* magnetic field,
which was due to a choice of degenerate eigenvalue. Now we evaluate
the current and density at finite magnetic field, to demonstrate
the intended effect.

Closes #150.

Rafal Skolasinski authored 7 years ago and Joseph Weston committed 7 years ago

A regression in Sympy causes symbols that are invalid
identifiers to be created, e.g. '-B_x'. While we cannot
fix Sympy's behaviour, we can at least shield our users
from its quirks.

Closes #179.

Previously, the wording of the explanatory comment above
discretized hopping functions was confusing. Now we
are clear about where the hopping originates.

Fixes #173.

Improve error message when creating invalid HoppingKinds

Closes #147, see !183.

The test was using a 2D delta with a 3D site family. This was
not previously raising an error because the HoppingKind in question
was not used for anything (just to test its equality to other
HoppingKinds).

Set correct stack level when issuing warnings, and when
calling API functionality that itself raises warnings.

See #161 and !181.

Some API functions call other API functions that report warnings.
We need to increase the stacklevel so that the user sees their
own line of code in the warning message.

Now the line in the *calling* code will be displayed, which is
more useful for users.

Closes #177.

This completes 9fdd41e8.

fix wraparound logic, add tests to fail cases and comments to clarify conventions of fundamental domains

apply position transformation before determining dimensionality;
this allows 2D projections of 3D systems to be plotted.

See merge request !161

This reverts commit 1e6c0860, reversing
changes made to 2c5da944.

The changes to Kwant's Lapack wrappers depend on features from
Scipy 0.16, so it cannot be merged into stable branch, which
depends on Scipy 0.14.

fix pickling on Cython 0.26

See merge request !175

Closes #170

enable PDF documentation builds

See merge request !171

Produce vector graphics, rather than rasterizing,
as rsvg-convert does.

Add the 'unicode-math' package and set the
latex engine to 'xelatex'.

Closes #126.

improve 'attach_lead' docstring

See merge request !170

sympy is installed in the Kwant testing image now.

We can now use Scipy's Cython interface to Lapack, which removes the need for us to explicitly link against blas or lapack ourselves. Also cut down boilerplate by re-arrange the existing Lapack wrappers to use Cython fused types.

Closes #28

See merge request !149