The High-Throughput Toolkit (httk)

The High-Throughput Toolkit (httk) is a toolkit for preparing and running calculations, analyzing the results, and storing results in global and/or personalized databases. httk is presently targeted at atomistic calculations in materials science and electronic structure, but aims to be extended into a library useful also outside those areas.

httk was created in 2014. This site describes httk₂, the current main version.

httk₂ is a rewrite of httk as a modular toolkit: instead of a single monolithic package, its functionality is split across independent module repositories that share a common, PEP 420 native httk.* namespace (httk.core, httk.io, httk.atomistic, and more). This lets you install and depend on only the parts you need, while httk.core provides the shared plugin, loading, and view/backend machinery the other modules build on.

Looking for httk version 1? The legacy site is available at /v1/.

Installation

The httk2 metapackage — which installs a sensible selection of modules in one step — is currently being set up, so the two commands below are not yet functional. For working, per-module installation instructions, see the full documentation at https://docs.httk.org.

Clone the source code repository and install it:
System Message: WARNING/2 (<string>, line 23)

Cannot analyze code. Pygments package not found.
```
.. code:: bash

   git clone https://github.com/httk/httk2.git
   pip install ./httk2
```
(not yet functional — see the note above)
Or, once released, install the metapackage from PyPI:
System Message: WARNING/2 (<string>, line 32)

Cannot analyze code. Pygments package not found.
```
.. code:: bash

   pip install httk2
```
(not yet functional — see the note above)

Quickstart

A few short general httk code examples follow in sections below.
Quickstarts covering specific functionalities are available for working with:

A few simple usage examples

Load a cif file

Loading a .cif file directly into an httk₂ Structure object is still under development — TBA. (httk₂ can already parse cif files into raw data via httk.core.load, but assembling a Structure from that parsed data is being ported.)

Create structures in code

In httk₂, a Structure is created from an explicit basis, a list of sites in reduced coordinates, a list of species (given as OPTIMADE-style species dictionaries), and a per-site list naming the species at each site. Here is a conventional cubic rock-salt (NaCl) cell:

System Message: WARNING/2 (<string>, line 63)

Cannot analyze code. Pygments package not found.

.. code:: python

  from httk.atomistic import Structure, StructurePrimitiveView

  basis = [[5.64, 0.0, 0.0],
           [0.0, 5.64, 0.0],
           [0.0, 0.0, 5.64]]

  sites = [[0.0, 0.0, 0.0], [0.5, 0.5, 0.0], [0.5, 0.0, 0.5], [0.0, 0.5, 0.5],
           [0.5, 0.5, 0.5], [0.0, 0.0, 0.5], [0.0, 0.5, 0.0], [0.5, 0.0, 0.0]]

  species = [
      {"name": "Na", "chemical_symbols": ["Na"], "concentration": [1.0]},
      {"name": "Cl", "chemical_symbols": ["Cl"], "concentration": [1.0]},
  ]

  species_at_sites = ["Na", "Na", "Na", "Na", "Cl", "Cl", "Cl", "Cl"]

  struct = Structure(basis=basis, sites=sites,
                     species=species, species_at_sites=species_at_sites)

  print("Species:", [s.name for s in struct.species])
  print("Number of sites:", len(struct.sites))

  # A view presents the same structure as a primitive (lattice, positions, numbers) triple:
  lattice, positions, numbers = StructurePrimitiveView(struct)
  print("Atomic numbers:", numbers)

Running this generates the output:

Species: ['Na', 'Cl']
Number of sites: 8
Atomic numbers: (11, 11, 11, 11, 17, 17, 17, 17)

Databases

Database storage and querying is still TBA — under development for httk₂.

Reporting bugs

Please file bugs at the issue tracker of the relevant module repository within the httk GitHub organization (please search first to check if it is already reported):

https://github.com/httk

Citing httk in scientific works

This is presently the preferred citation:

R. Armiento et al., The High-Throughput Toolkit (httk), http://httk.org/; Armiento R. (2020) Database-Driven High-Throughput Calculations and Machine Learning Models for Materials Design. In: Schütt K., Chmiela S., von Lilienfeld O., Tkatchenko A., Tsuda K., Müller KR. (eds) Machine Learning Meets Quantum Physics. Lecture Notes in Physics, vol 968. Springer, Cham. https://doi.org/10.1007/978-3-030-40245-7_17

Since httk may call upon many other pieces of software quite transparently, it may not be initially obvious what other software should be cited. Unless configured otherwise, httk prints out a list of citations when the program ends. You should take note of those citations and include them in your publications if relevant.