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.


  • If you just want to use httk, this should work:

    pip install httk
    • If your pip works as it should, you should be able to do import httk in Python, as well as run the httk command line tool.
  • An alternative is to clone the master branch of our source code repository, which is meant to always give you the latest release:

    git clone

    Then, every time you want to use httk you need to initialize the environtment with:

    source /path/to/httk/

    (Which, of course, can be put in your shell initialization scripts)

  • For more installation alternatives, see the full documentation.


A few simple usage examples

Load a cif file or poscar

This is a very simple example of just loading a structure from a .cif file and writing out some information about it.

import httk

struct = httk.load("example.cif")

print("Formula:", struct.formula)
print("Volume:", float(struct.uc_volume))
print("Assignments:", struct.uc_formula_symbols)
print("Counts:", struct.uc_counts )
print("Coords:", struct.uc_reduced_coords)

Running this generates the output:

('Formula:', 'BO2Tl')
('Volume', 509.24213999999984)
('Assignments',['B', 'O', 'Tl'])
('Counts:', [8, 16, 8])
('Coords', FracVector(((1350,4550,4250) , ... , ,10000)))

Create structures in code

from httk.atomistic import Structure

cell = [[1.0, 0.0, 0.0] ,
        [0.0, 1.0, 0.0] ,
        [0.0, 0.0, 1.0]]
coordgroups = [[
                  [0.5, 0.5, 0.5]
                  [0.0, 0.0, 0.0]
                  [0.5, 0.0, 0.0], [0.0, 0.5, 0.0], [0.0, 0.0, 0.5]

assignments = ['Pb' ,'Ti' ,'O']
volume =62.79
struct = Structure.create(uc_cell = cell,
             uc_reduced_coordgroups = coordgroups,
             assignments = assignments,
             uc_volume = volume)

Create database file, store a structure in it, and retrive it

import httk, httk.db
from httk.atomistic import Structure

backend = httk.db.backend.Sqlite('example.sqlite')
store =

tablesalt = httk.load('NaCl.cif')

arsenic = httk.load('As.cif')

# Search for anything with Na
search = store.searcher()
search_struct = search.variable(Structure)

search.output(search_struct, 'structure')

for match, header in list(search):
    struct = match[0]
    print "Found structure", struct.formula, [str(struct.get_tags()[x]) for x in struct.get_tags()]

Create database file and store your own data in it

#!/usr/bin/env python

import httk, httk.db
from httk.atomistic import Structure

class StructureIsEdible(httk.HttkObject):

    @httk.httk_typed_init({'structure': Structure, 'is_edible': bool})
    def __init__(self, structure, is_edible):
        self.structure = structure
        self.is_edible = is_edible

backend = httk.db.backend.Sqlite('example.sqlite')
store =

tablesalt = httk.load('NaCl.cif')
edible = StructureIsEdible(tablesalt, True)

arsenic = httk.load('As.cif')
edible = StructureIsEdible(arsenic, False)

Reporting bugs

Please file bugs at the issue tracker at github (please search first to check if it is already reported):

Citing httk in scientific works

This is presently the preferred citation:

  • R. Armiento et al., The High-Throughput Toolkit (httk),; 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.

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.

More documentation

More extensive documentation about httk is available at