Regular Online Meetings
- Two final PRs before release: magnetics
flux_loopof differential type (https://github.com/iterorganization/IMAS-Data-Dictionary/pull/137) and rationalisation of GGD structures between edge* and plasma* IDSs (https://github.com/iterorganization/IMAS-Data-Dictionary/pull/150 --> one test is failing and requires fix) - Identifiers in this version introduce the possibility of keeping old names as alias after a renaming, for backward compatibility (AL has been patched to handle this capability)
-
summaryin this version was extended to better track important metadata in simdb, which is now in production at ITER: https://simdb.iter.org
- Done with EFIT++ on a few test datasets kindly made available for this purpose: works for HL3, issues spotted on AUG (old structure for storing
Ip) and with MAST-U (coils/loop with no current, error in EFIT++ solver)"In the case of PC I would recommend constraining it to 0 current because the coil exists but is not connected. If you set the weight to zero then EFIT will be free to put current there when there is actually none "
- Advise to use issues in this github repo to keep track of potential issues and fixes
- During the general assembly, EF agreed to have IO releasing key software that include their BIP under LGPL (imas-core, idstools, access-layer APIs)
- ets-core is already released under EUPL (could be extended to a more permissive one if needed)
- preparation work had started for imas-core, towards a distribution package made available in PyPI that contains both HDF5 and UDA backend (Linux, only HDF5 in Windows package for the moment, and none with MDSplus which has difficulties building and at runtime)
- DMP (Data Management Plan) hope to pursue scenario D with a few full discharges mapped and made available, and open-access data is expected (after embargo period) for both WEST and TCV. The timescale for this project is 2 years, but it will benefit extra fundings.
- Please forward to interested parties --> subscribe on INDICO
- Working session to test/fix example data, aim at unifying mapping syntax and sharing good practices and tips
- Discuss and decide on top priorities + use-cases (modelling and analysis) for extending the mappings
Quick overview of the GitHub repository
- A few PR require work before merging.
- Additional information could be hosted in separated folders in this repository and linked from the top-level table (links can also lead to external pages if these are publicly accessible).
- Examples of open-access IDS data for ITER are being hosted on Zenodo for part of the machine description and a few simulations.
- Registrations are open in INDICO.
- Please register before 31 October and forward the link to relevant people.
- @DavidPCoster presented the current status as gathered by the DMP (Data Management Plan in EUROfusion) about the available mappings of various European tokamaks.
- Here available means that there are existing tools and mappings that can be applied to potentially any pulses.
- This is a living DMP internal document, could serve to help structure the information in this repository (DD version, raw vs reconstructed signals, suit of codes --e.g. ETS or IDA-- that can make use of the various IDS data...).
- Some mappings are the result of analysis codes, e.g. TRANSP, it will be interesting to know which version of the code.
- @ApiwatWis introduced briefly the current effort to try mapping and using IMAS on TT-1 in Thailand.
- He Mentioned issues to install some dependencies of IMAS software locally, which needs to be described in the relevant repositories or any other channels such as the IMAS team can advise.
- @DavidPCoster pointed to this discussion about which IDS to use to store some diagnostic data, and the initial answers given with ChatGPT 5 and the IMAS MCP server by @Simon-McIntosh can be found in Slack and below.
* You can find more information about this MCP server at [https://github.com/iterorganization/imas-mcp](https://github.com/iterorganization/imas-mcp). Don't hesitate to try it and give feedback about the relevance of the answers with such tool.
- IMAS-Data-Dictionary 4.1.0 (coming 1-2 weeks)
- Will involve updates of the identifier libraries in some languages.
- AL-Core 5.5.0
- Shall contain the fix for making UDA compatible with complex numbers (requires also a new upcoming release of the IMAS UDA-plugins as well as a release of UDA-core 2.8.2 which is pending)
- Fix nocache mode which will make the UDA backend fully compatible with IMAS-Python's lazy loading
- SimDB/Dashbaord soon in production at ITER