AiiDA
Automated Interactive Infrastructure and Database for Computational Science


If you use AiiDA please cite:
AiiDA 1.x
Main paper: S.P. Huber et al., Scientific Data 7, 300 (2020)
AiiDA engine: M. Uhrin et al., Comp. Mat. Sci. 187 (2021)

AiiDA 0.x
First paper, ADES model: G. Pizzi et al. Comp. Mat. Sci. 111, 218-230 (2016) (open access version)

Posts

Newly-engineered Materials Cloud Archive

👤 🕔 May 27, 2020 Comments Off on Newly-engineered Materials Cloud Archive

We would like to announce the launch of a newly engineered Materials Cloud Archive, now powered by the same Invenio framework as the massive Zenodo repository at CERN.

Read more…

2020 Questionnaire results – AiiDA papers & testimonials

👤 🕔 May 9, 2020 Comments Off on 2020 Questionnaire results – AiiDA papers & testimonials

The results of the annual questionnaire on AiiDA-powered research projects are out!

Find them on the new aiida.net/science page where we plan to collect future survey results and keep track of publications using AiiDA.

We’ve also asked AiiDA users what they use AiiDA for and how it benefits their day-to-day research. Find their take on aiida.net/testimonials.

AiiDA v1.2.0 released

👤 🕔 April 15, 2020 Comments Off on AiiDA v1.2.0 released

A new AiiDA release v1.2.0 is available! You can find more information at our download page. It can be installed through pip as:

    pip install aiida-core==1.2.0
    reentry scan

 

N.B.:  the `reentry scan` step is crucial as we have added new entry points.

This is the second minor version of the v1 series and comes with mostly new features.
Especially groups have received a lot of new functionality.

Groups can now be subclassed
The `Group` class can now be subclassed to implement custom functionality, just like for example the `Data` node class. They can be registered through entry points in the new entry point category “aiida.groups”. Please refer to the documentation [1] for more details.

Virtual group hierarchy through `GroupPath`
The group concept in AiiDA so far has been “flat”. That is to say, group can contain nodes, but you a group cannot contain another group. Many users have expressed the wish to be able to treat groups a bit more like folders on a filesystem and add some hierarchy to them. Thanks to a contribution by Chris Sewell, the `GroupPath` utility now provides a virtual hierarchy based on the group labels. Forward slash characters ‘/’ in group labels will be interpreted as sub groups. A very short demonstration:

In [1]: from aiida.tools.groups import GroupPath
   ...: Group(label='project/sub/a')
   ...: Group(label='project/sub/b')

In [2]: path = GroupPath('project/sub') 
   ...: for child in path.children: 
   ...:     group = child.get_group() 
   ...:     print('Group<{}> contains {} nodes'.format(group.label, group.count())) 
   ...:
Group<project/sub/a> contains 0 nodes
Group<project/sub/b> contains 0 nodes

Of course the `GroupPath` class provides much more than that and it is also exposed on the CLI through `verdi group path ls`. Extensive documentation of all functionality will be added to the official documentation soon.

More details and download options can be found through the following links:

Pre-prints of upcoming AiiDA & Materials Cloud papers now available

👤 🕔 March 30, 2020 Comments Off on Pre-prints of upcoming AiiDA & Materials Cloud papers now available

After five years of continued development since the first AiiDA paper it was time to sit down and recap how AiiDA has evolved. You can read a pre-print of the upcoming paper on AiiDA 1.0 at Huber, S.P. et al., arXiv:2003.12476 (2020)

We would like to take this opportunity to acknowledge and thank all those who have contributed to this milestone with code, bug fixes, improvements of the documentation, and useful discussions and suggestions:

Oscar D. Arbeláez-Echeverri, Michael Atambo, Valentin Bersier, Marco Borelli, Jocelyn Boullier, Jens Bröder, Ivano E. Castelli, Keija Cui, Vladimir Dikan, Marco Dorigo, Y.-W. Fang, Espen Flage-Larsen, Marco Gibertini, Daniel Hollas, Eric Hontz, Jianxing Huang, Christoph Koch, Ian Lee, Daniel Marchand, Antimo Marrazzo, Simon Pintarelli, Gianluca Prandini, Philipp Rüßmann, Philippe Schwaller, Ole Schütt, Christopher Sewell, Andreas Stamminger, Atsushi Togo, Daniele Tomerini, Nicola Varini, Jason Yu, Austin Zadoks Bonan Zhu, Mario Zic, as well as all coauthors of the manuscript.

We have also just uploaded a pre-print describing the Materials Cloud platform powered by AiiDA in case you’re interested: Talirz, L. et al., arXiv:2003.12510 (2020)

The AiiDA Team

P.S. If your scientific work has benefited from the use of AiiDA 1.0 or the Materials Cloud, please acknowledge this by citing the pre-prints (until the papers are out).

AiiDA at Google Summer of Code 2020

👤 🕔 March 5, 2020 Comments Off on AiiDA at Google Summer of Code 2020

Thanks to the folks at NumFOCUS, AiiDA is participating in the Google Summer of Code 2020 under the NumFOCUS umbrella for the first time.

March 5th – Call for participation 

In short, this means that students (at the BSc/MSc or Ph.D. level, see eligibility criteria) can apply to work on an interesting AiiDA coding project for 3 months during the summer (May 18th-August 10th) and get reimbursed for their work.

Applications officially start March 16th, but selection will be competitive, and we recommend

Looking forward to your applications! We’ll keep updating this post as we move along.

The AiiDA Team

P.S. We are committed to help fight the insane gender imbalance in the open source community – that is to say: women are particularly encouraged to apply.

May 5th – GSOC students announced

We’re excited to announce that Jason Eu, a Ph.D. student from the Department of Physics at South China University of Technology, China will be participating in this GSOC working on adding asyncio support in plumpy, aiida-core and circus!

Welcome to the team, Jason! You can follow Jason’s progress on his blog.

AiiDA v1.1.1 released

👤 🕔 March 3, 2020 Comments Off on AiiDA v1.1.1 released

A new AiiDA release v1.1.1 is available! You can find more information at our download page. It can be installed through pip as:

    pip install aiida-core==1.1.1

This is the first patch release of the v1.1 series and as such contains mostly bug fixes and some minor changes and improvements. The most important changes relate to the caching mechanism and the BaseRestartWorkChain.

Caching mechanism
Due to a bug, certain calculations were not cached where they should, which has now been fixed. In addition, the format of the caching configuration file was broken, which has now been fixed (see the documentation).

BaseRestartWorkChain
The `BaseRestartWorkChain`class and associated utilities were added for beta trial in `aiida-core==1.1.0`. We have had useful discussions with developers and users during the AiiDA hackathon at CINECA from 17 – 21 February (see the report).

As a result of these discussions, we have decided to change the use of the `register_process_handler`. It has been renamed to `process_handler` and can now only be used to decorate instance methods of the work chain class within its scope and no longer outside of it. In addition, the decorator has a new keyword argument `enabled`, which determines whether the handler is by default considered during process handling. This default value can be overridden on a per work chain basis through the new `handler_overrides` input. Since the interface changed since its introduction in `aiida-core==1.1.0`, please only use the `BaseRestartWorkChain` from `aiida-core>=v1.1.1`.

More details and download options can be found through the following links:

Notes from AiiDA hackathon on plugin and workflow development

👤 🕔 March 3, 2020 Comments Off on Notes from AiiDA hackathon on plugin and workflow development

The  AiiDA hackathon held at CINECA from February 17th-21st 2020 featured a number of presentations and discussion sessions on topics relevant to plugin and workflow development.

Those interested can find links to slides of presentations, video recordings, as well as notes of the discussion sessions in the report . Enjoy!

AiiDA v1.1.0 released

👤 🕔 February 12, 2020 Comments Off on AiiDA v1.1.0 released

A new AiiDA release v1.1.0 is available! You can find more information at our download page.

This is the first minor release of the v1 series, in which we finally drop the support for older python 2 versions (critical bug fixes for python 2 will be supported until July 1 2020 on the `v1.0.*` release series) and add support for python 3.8, thus becoming compatible with all current python versions that are not end-of-life (i.e.: 3.5, 3.6, 3.7, 3.8). You can find the reasoning for doing this without a major version change in the corresponding AEP 001.

The update also includes some bug fixes and performance improvements, as well as some interesting new features such us as new tools for traversing the AiiDA provenance graphs (AiiDA Graph Explorer), which is now used for consistently deleting and exporting nodes with a considerable increase in performance, and tools for automated error handling for workchains (BaseRestartWorkChain). Finally, great efforts have been made towards relaxing the dependencies of the code in order to make it easier to install AiiDA side-by-side with other packages: plugin developers are encouraged to adopt a similar philosophy, going through their own dependencies and relaxing them where applicable.

More details and download options can be found through the following links:

Change log

Clone url

Download zip

AiiDA becomes NumFOCUS affiliated project

👤 🕔 February 5, 2020 Comments Off on AiiDA becomes NumFOCUS affiliated project

We are excited to announce that AiiDA has joined the ranks of NumFOCUS affiliated projects !

NumFOCUS is a public charity that promotes open practices in research, data, and scientific computing. It provides an extensive network of great open source projects in the computational science domain, and the AiiDA team is planning to participate in the Google Summer of Code under the NumFOCUS umbrella. We’ll keep you posted!

AiiDA Hackathon: developing code plugins and robust scientific workflows @ Cineca, Bologna, Italy

👤 🕔 January 22, 2020 Comments Off on AiiDA Hackathon: developing code plugins and robust scientific workflows @ Cineca, Bologna, Italy

The hackathon will take place from February 17  to 21 at Cineca: Via Magnanelli, 6/3, 40033 Casalecchio di Reno, Italy.
Participating AiiDA plugin developers will be shown best practices and useful tips on:

  •  code development practices (automated testing, continuous integration, code style checks, packaging, distribution)
  • writing reusable, robust and modular workflows
  • taking full advantage of the AiiDA v1.0 API

In addition, during the event we will discuss and define common APIs for workflows for the computation of materials’ properties.

When adopted by plugin developers, these common APIs enable AiiDA users to compute a material’s property using different codes without the need to know the interface of each plugin in detail.

Program

  • The workshop will start on February 17th at 15:00 and will finish on February 21st at 12:00
  • Welcome and registration on 17th will be open from 14:30
  • The current tentative program can be found on the PRACE website

Target audience
Computational scientists that already actively or plan to develop/maintain one or more AiiDA plugins (registered on the AiiDA plugin registry), in order to automate calculations with AiiDA.

Prerequisites
Being experienced with using AiiDA in general and familiar with developing calculation plugins and workflows

Registration
* Link: https://events.prace-ri.eu/event/957/overview
* Deadline: February 1st

Financial
* Subscription is free
* Travel costs may be partially reimbursed for qualifying applicants whose institution cannot reimburse.
For questions contact Francisco Ramirez (francisco.ramirez@epfl.ch) or Sebastiaan Huber (sebastiaan.huber@epfl.ch)

Organizers
Nicola Spallanzani, Giovanni Pizzi, Sebastiaan Huber, Francisco Ramirez, Miki Bonacci, Emanuele Bosoni, Vasily Tseplyaev, Fabio Affinito.

Acknowledgements

We acknowledge financial support by PRACE and by the EU Centre of Excellence MaX “MAterials design at the eXascale” (a H2020-INFRAEDI-2018-1 funded project Grant Agreement n. 824143) and by NCCR MARVEL funded by the Swiss National Science Foundation.