v0.9.1 | 2017-04-28¶
- Fixed an issue with multi-orbital models where onsite/hopping modifiers would return unexpected
results if a new
energyarray was returned (rather than being modified in place).
Solver.calc_probabilityreturning single-orbital results for multi-orbital models.
- Fixed slicing of
Structureobjects and made access to the
v0.9.0 | 2017-04-14¶
- This version includes extensive internal improvements and raises the minimum requirements for installation. Starting with this release, only Python >= 3.5 is supported. Newer versions of the scientific Python packages are also required: numpy >= 1.12, scipy >= 0.19 and matplotlib >= 2.0.
- On Linux, the minimum compiler requirements have also been increased to get access to C++14 for the core of the library. To compile from source, you’ll need GCC >= 5.0 or clang >= 3.5.
- Improved support for models with multiple orbitals, spins and any additional degrees of freedom. These can now be specified simply by inputing a matrix as the onsite or hopping term (instead of a scalar value). For more details, see the “Multi-orbital models” section of the documentation.
- Lifted all limits on the number of sublattices and hoppings which can be defined in a
Latticeobject. The previous version was limited to a maximum of 128 onsite and hopping terms per unit cell (but those could be repeated an unlimited number of times to form a complete system). All restrictions are now removed so that the unit cell size is only limited by available memory. In addition, the memory usage of the internal system format has been reduced.
- Added a 3-band model of group 6 transition metal dichalcogenides to the Material Repository. The available TMDs include: MoS2, WS2, MoSe2, WSe2, MoTe2, WTe2. These are all monolayers.
- Complicated system geometries can now be created easily by composing multiple simple shapes. This is done using set operations, e.g. unions, intersections, etc. A complete guide for this functionality is available in the “Composite shapes” section of the documentation.
Kernel polynomial method¶
- The KPM implementation has been revised and significantly expanded. A guide and several examples
are available in the “Kernel polynomial method” section of the documentation (part 9 of the
Tutorial). For a complete overview of the available methods and kernels, see the
chebyshevsection of the API reference.
- New builtin computation methods include the stochastically-evaluated density of states (DOS) and electrical conductivity (using the Kubo-Bastin approach).
- The new low-level interface produces KPM expansion moments which allows users to create their own KPM-based computation routines.
- The performance of various KPM computations has been significantly improved for CPUs with AVX support (~1.5x speedup on average, but also up to 2x in some cases with complex numbers).
- Added the
pb.load()convenience functions for getting result objects into/out of files. The data is saved in a compressed binary format (Python’s builtin
pickleformat with protocol 4 and gzip). Loaded files can be immediately plotted:
result = pb.load("file.pbz")and then
result.plot()to see the data.
- The eigenvalue solvers now have a
calc_ldosmethod for computing the local density of states as a function of energy (in addition to the existing
- Improved plotting of
Latticeobjects. The view can now be rotated by passing the
axis="xz"argument, or any other combination of x, y and z to define the plotting plane.
Deprecations and breaking changes¶
Lattice.add_aliases()method. The old
Lattice.add_sublattice(..., alias=name)way of creating aliases is deprecated.
greensmodule has been deprecated. This functionality is now covered by the KPM methods.
- The internal storage format of the
Systemclasses has been revised. This shouldn’t affect most users who don’t need access to the low-level data.
v0.8.2 | 2017-01-26¶
- Added support for Python 3.6 (pybinding is available as a binary wheel for Windows and macOS).
- Fixed compatibility with matplotlib v2.0.
- Fixed a few minor bugs.
v0.8.1 | 2016-11-11¶
- Structure plotting functions have been improved with better automatic scaling of lattice site circle sizes and hopping line widths.
- Fixed Brillouin zone calculation for cases where the angle between lattice vectors is obtuse (#1). Thanks to @obgeneralao (Oliver B Generalao) for reporting the issue.
- Fixed a flaw in the example of a phosphorene lattice (there were extraneous t5 hoppings). Thanks to Longlong Li for pointing this out.
- Fixed missing CUDA source files in PyPI sdist package.
- Revised advanced installation instructions: compiling from source code and development.
v0.8.0 | 2016-07-01¶
- Added support for scattering models. Semi-infinite leads can be attached to a finite-sized scattering region. Take a look at the documentation, specifically section 10 of the “Basic Tutorial”, for details on how to construct such models.
- Added compatibility with Kwant for transport calculations. A model
can be constructed in pybinding and then exported using the
Model.tokwant()method. This makes it possible to use Kwant’s excellent solver for transport problems. While Kwant does have its own model builder, pybinding is much faster in this regard: by two orders of magnitude, see the “Benchmarks” page in the documentation for a performance comparison.
- Experimental: Initial CUDA implementation of KPM Green’s function (only for diagonal elements for now). See the “Experimental Features” section of the documentation.
- The performance of the KPM Green’s function implementation has been improved significantly: by a factor of 2.5x. The speedup was achieved with CPU code using portable SIMD intrinsics thanks to libsimdpp.
- The Green’s function can now be computed for multiple indices simultaneously.
- The spatial origin of a lattice can be adjusted using the
Lattice.offsetattribute. See the “Advanced Topics” section.
- The interface for structure plotting (as used in
StructureMap) has been greatly improved. Some of the changes are not backwards compatible and may require some minor code changes after upgrading. See the “Plotting Guide” section of the documentation for details.
- The interfaces for the
StructureMapresult objects have been revised. Specifically, structure maps are now more consistent with ndarrays, so the old
smap.filter(smap.x > 0)is replaced by
smap2 = smap[smap.x > 0]. The “Plotting Guide” has a few examples and there is a full method listing in the “API Reference” section.
- The API reference has been completely revised and now includes a summary on the main page.
- A few advanced topics are now covered, including some aspects of plotting. A few more random examples have also been added.
- Experimental features are now documented.
- Fixed translational symmetry skipping directions for some 2D systems.
- Fixed computation of off-diagonal Green’s function elements with
opt_level > 0
- Fixed some issues with shapes which were not centered at
(x, y) = (0, 0).
v0.7.2 | 2016-03-14¶
- Lots of improvements to the documentation. The tutorial pages can now be downloaded and run interactively as Jupyter notebooks. The entire user guide is also available as a PDF file.
hop_idmodifier arguments can now be compared directly with their friendly string names. For example, this makes it possible to write
sub_id == 'A'instead of the old
sub_id == lattice['A']and
hop_id == 'gamma1'instead of
hop_id == lattice('gamma1').
- The site state modifier can automatically remove dangling sites which have less than a certain
number of neighbors (set using the
- Added optional
sitesargument for state, position, and onsite energy modifiers. It can be used instead of the
x, y, z, sub_idarguments and contains a few helper methods. See the modifier API reference for more information.
- Fixed a bug where using a single KPM object for multiple calculations could return wrong results.
hopping_generatorwhich can be used to add a new hopping family connecting arbitrary sites independent of the main
Latticedefinition. This is useful for creating additional local hoppings, e.g. to model defects.
v0.7.1 | 2016-02-08¶
Added support for double-precision floating point. Single precision is used by default, but it will be switched automatically to double if required by an onsite or hopping modifier.
Added support for the 32-bit version of Python
Tests are now included in the installed package. They can be run with:
import pybinding as pb pb.tests()
Available as a binary wheel for 32-bit and 64-bit Windows (Python 3.5 only) and OS X (Python 3.4 and 3.5)
v0.7.0 | 2016-02-01¶