DIRAC

FKF Meeting Room (FKF Møderum Ø10-412-2)

To get the SDU map link to the meeting room, type FKF in the search field at this link: SDU Møderum

The DIRAC meeting is a family reunion for DIRAC developers and DIRAC friends. It is a nice opportunity to meet the people behind the DIRAC code. There is no registration fee but for organization purposes please add yourself to the participant list below (send an e-mail to Hans Jørgen using hjj at sdu.dk if you do not have editing permission to this page).

Most people will stay at the campground (more info below). The cabins are very good comfort, not expensive, and it's fun to be there. Book them well in advance.

We will discuss science but a significant part of the program is reserved to discuss the future of the DIRAC code, programming, code maintenance, release planning, as well as for programming task groups.

The meeting program is self-organised, simply add talks and discussion points to the program.

Typically you will fly to Copenhagen and then take the train or fly to Billund and then take bus to Vejle plus train Vejle-Odense. From Odense train station you can take bus 61 or 62 (in the right direction!) or share cab to the DCU-Camping Odense. If you arrive to Odense by train on a regular train ticket, then the bus from the railway station to the camp ground (or any other destination in Odense) is included in the train ticket.

Make your cabin reservation at DCU-Camping Odense. The cabins can accommodate 3-4 people so organize with others. Use the dirac-developers mailing list.

A downtown hotel at the railway station: CABINN Odense Hotel. You can find other hotels from around 500 DKK per night by searching google for “hotels Odense” and some BBs to less than 500 DKK. Note that hotels might be cheaper via their own web site than on hotels.com etc.

Due to the Rammstein concerts on Friday and Saturday, June 2 and 3, Odense is busier than usual: book in time !

• Fynbus web page is here. Notice that there is a Fynbus app.
• Regular bus tickets for local travel in Odense cost DKK 24 in cash for a single ride (including changing lines) and are valid for 1 hour.
• With the fynbus app it is the same price, but you will be able to pay with your credit card. The ticket must be bought before entering the bus and you show the ticket on your smart phone when you enter the bus.
• Bus routes Railway station - campground (line 61-62)
• Bus routes Railway station - university (line 41-44)
• A map of all bus lines in Odense can be found here
• Arriving by train to Odense on a regular train ticket (not discount “orange ticket” bought on http://www.dsb.dk/en/) you can use your train ticket for the bus and get a free ride (to the campground).
• (Not recommended by Hans Jørgen, unless you plan to travel in Denmark regularly.) You can buy a public transport card (“Rejsekort anonymous”) valid in entire Denmark already at the airport in Copenhagen which you can charge and use for all train and bus rides. However it is quite expensive in start-up fee (110 DKK) + 600 DKK pre-paid to be able to use it for national travel (e.g. to CPH-Odense). In general, you have to charge it with enough money in a “Rejsekort automat” for any travel before the travel. As far as we can see you cannot get a refund of unused money. Furthermore, without a Danish address you won't be able to get receipt for the different travelled distances.http://www.rejsekort.dk

Please note arrival and departure if you will not participate the whole time.

1. Hans Jørgen Aa. Jensen, Odense
2. Trond Saue, Toulouse
3. Ayaki Sunaga, Kyoto (online, if possible)
4. Martin van Horn, Toulouse
5. Gabriele Fabbro, Toulouse
6. Luuk Visscher, Amsterdam
7. Andre Gomes, Lille
8. Xiang Yuan, Lille
9. Erik Hedegård, Odense
10. Ernst D. Larsson, Odense
11. Jan Brandejs, Toulouse
12. Johann Pototschnig, Toulouse
13. Gosia Olejniczak, Warsaw

—-

• Discuss simplified installation using Conda (or similar, e.g. Spack).
• Go through issues list
• Document (some of the) available DFT functionals, incl. keyword (question on dirac-users)
• Test .HTSORT option, if not fixed before meeting (dirac-users: MEMGET error)
• Experience with HDF5 files: one-electron integrals too voluminous - revised SCHEMA for D24 ?
• All executables link to all object files. Is this really needed ?
• Standard data types for: (1) orbital information, (2) ao basis set, (3) symmetry info.
• Prepare (agree on) the basic structure of the unified tensor interface and draft the tensor type
• Installation on supercomputers: LUMI / Frontier session?

• Arrival
• 19:30 Oriental Barbecue House (Slotsgade 20) - booked for 10 persons (LV)

• 09:00 – 09:15 Opening, discussion of meeting program (Hans Jørgen)
• 09:15 – 09:30 Introduction round (everybody)
• 09:30 – 11:30 Overview of current and planned DIRAC developments by node leaders
• 11:30 – 12:00 Task group formation

• 12:00 – 13:30

• 13:30 – 14:00 git commits should compile (Trond)
• 14:00 – 14.30 unified tensor interface in DIRAC (Jan)
• 14:30 – 15.15 checkpointing in exacorr ? (Andre)
• 15:15 – 15:45

• 15.45 – 17.30 Online Session: Focus on coupled cluster development.

• 09:00 – 09:30 Analysis of electric field gradient in molecules in relativistic framework (Gabriele)
• 09:30 – 10:00 A story about localizing a localization bug (Trond)
• 10:00 – 11:00 coffee break and group picture
• 11:00 – 11:30 MCSCF without symmetry: a status report (Martin)
• 11:30 – 12:00 Generating CC code for modern distributed memory tensor software (Jan)

• 12:10 – 13:30

• 13:30 – 14:00 Integral libraries and interfaces (Johann)
• 14:00 – 14:30 Integral requirements for non-perturbative treatment of magnetism (Ansgar)
• 14:30 – 15:30 Discussion

• 16:00 – 17:30

• 19:00 Dinner at Storms Pakhus

• 09:00 – 09:30 ExaCorr updates: EOM and CC response property (Xiang Yuan)
• 09:30 – 10:00 Polarizable embedding with relativistic wave functions (Erik)
• 10:30 – 11.00 Localizing the wave function, why and how ? (Luuk, optional)
• 11:00 – 12:30 Collaborative work and coffee

• 13:30 – 14:00 Revision of data Schema for DIRAC24
• 14:00 – 14:30 Checkpointing and schema of real-space data + updates in VISUAL (Gosia, notes)
• 14:30 – 15:00 Quaternion angular functions (Trond;optional)

• 15:00 – 17:30

• 09:00 – 17:30 Collaborative session (joint papers, joint programming efforts).