Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
π½ β¬οΈ Upgrade process to open source version (#2522)
* π§ First attempt at updating process to open source version * π½οΈ Data removal/moved * π½οΈ Non existant filepath changed * π¨ Add back install script * π¨ New url * π¨ Main * π Working env and process for OS version * π¨ Update logflush * β Fix tests * π¨ Comment out thermalshields * π Resolve conflict * WIP: Python API to PROCESS (#2616) * π½οΈ Data removal/moved * π½οΈ Non existant filepath changed * add NLopt version correction to install-process.sh * fix PROCESS radial build plot * fix PROCESS plotting better * first pass PROCESS inputs * first pass PROCESS inputs * formatting * finish constraint equations * iteration variables * iteration variables * iteration variables * start on bounds * to_invariable * bounds * fix constraints and change input values * start adding starting point and change dr_tk_cas etc * add fatigue variables * fix PROCESS to run but problems with pulse length and BB pumping power * slight progress * some cleanup * still not there * PROCESS documentation no longer has teh descriptions of ixc and icc RIP * PROCESS documentation no longer has teh descriptions of ixc and icc RIP * PROCESS documentation no longer has teh descriptions of ixc and icc RIP * minor * itvar numbers * pause * constraint equation mapping * objective equation mapping * objective equation mapping * duplication defense * favlue mapping * output * example * fuller example * add missing plasma models * missed one * add TF coil models * more switches * add eq var mapping file * move template builder to api.py * put template builder in separate file * split out model mappings * PROCESSModel * self not cls * flake8 flag for hack * flake8 flag for hack * docstring * incorporate mechanics into template builder * need to restart PC * start piping into example * slow an dsteady progress * pause before rebase * match install script * CHEEKY NOQA * whoops * typos and input additions * typos and input additions * more missing inputs * impurity handling * more or less done with the PROCESS API * inputs and tk_ts duplicaiton * string bound weirdness * remove spurious int * add missing fimp * fix missing primary pumping model * add mode bloody switches * Constraint enum * objective enum * pipe through and clean up * update radial build example * fix typing * tuple --------- Co-authored-by: james <james.cook1@ukaea.uk> * PROCESS API enhancement: required input values listed for model selections (#2659) * add model selection mechanics * do one and move to example * add template builder mechanics * concision * undo test mods * also check variables * populate some cases * minor * minor * start on constraint equation fancy enums * why do i have to do this * its the weekend and i am CS fatigue * fold in constraints * tuples not lists * finish constraints but a number of inconsistencies in PROCESS documentation * adjust_variable * warning fixes and catch typos and bugs * more honing in on IN.DAT * cpttf_mak removed * model requires_values tuple * better tuples of one * missing inputs for constraints * more missing inputs - not making this easy.. * remove ffuspow * add taulimit * taulimit * missing tuple of one * missing tuple of one part 2 * picky ruff * add test for bad var name * add test for repeated variable * add another warning test * check it was added too * check bad name adjust * test warn on overwrite * test impurity shenanigans * test warn on missing input model * test warn on overwrite model * was missing an assert * check for missing itvars * better check of existence for constraint itv warning * tuples of one * 24 is not used wtf * 55 not used * missing inputs * more not used * oacddp no longer used?! * missing fvale mapping * some sanity stuff but still not working * the longest it ever took to fail * missed a tuple of one * fix install script and add tests * match up EUDEMO run * update align some defaults * get rid of obsolete PROCESS names * adding endless names * remove wrong / obsolete names * the journey to nowhere continues * more names * divertor variables * PF stuff * physics * physics * TF reshuffle * final ordering clean up * final clean * final clean * couple minor tweaks * final tweak * explicitly apply default bounds to variables * typo * all bounds * start on None * nums to None * lists to None * ixc icc bounds None * default bounds * remove non default models * truly default bounds * constraint ordering * defaults * full defaults from ref IN.DAT * working but feels fragile * work on integration test * work on integration test * nearly there * remove fpinj as default f-value * test inclusion * value tweaks in raial build and runs * ποΈ Fixing PROCESS InDat writer (#2690) * ποΈ Fixing PROCESS InDat writer * π Need to import Parameter for dynamic typechecking * β Add test for template writing * β Test from params too * π¨ Add option to get params out * β Fix template builder tests * π Fix params that were not sent * π¨ Use self as it is self * π§ WIP None avoidance * π¨ Add back params * β»οΈ Cleanup * β Fix impurities test * β Fix tests * π Skip tests if process not installed * π Upgrade env and fix test * π¨ Update env * π Fix env * π Improve documentation * π Fix runtitle * π·οΈ Optional * π¨ Add runtitle * ποΈ Fix impurity data reading * π In eV not keV * π Docs * π¨ Use old env * π ClassVar breaks stuff * π Bad rebase * π¨ Fix formatting * π§ TMP environment changes * π Fix radial build * π Sort PROCESS iteration variables * Add PROCESS API example (#2740) * minor * complete example CBA to make a converging one that isnt EU-DEMO * preamble * add example to TOC t ree * update impurities id * udpate example * improve example * update fimp( to fimp() * π· Exclude process example from examples CI --------- Co-authored-by: james <james.cook1@ukaea.uk> --------- Co-authored-by: je-cook <81617086+je-cook@users.noreply.github.com> Co-authored-by: james <james.cook1@ukaea.uk> * Update PROCESS v3 (OS) mappings for EU-DEMO (#2749) * first pass mappings * long_name not long name * populate PROCESS params * unit issue * finish process params descriptions * π© Enable really bad units * mapping fixes * start on special ase mechanics * shield + vv shenanigans * finish applying hacks * add q_0 to test params * add mslimit to test data * add sigma cs * add missing params to test data * move q out of OUT mapping * add proper TODOO * β¨ Allow different in and out names for external codes (#2750) * β¨ Allow different in and out names for external codes * β Add tests * π§ A horrible hack for value=None unit=crap * π» Actually it might be ok * manually turn shldith send back on --------- Co-authored-by: james <james.cook1@ukaea.uk> Co-authored-by: je-cook <81617086+je-cook@users.noreply.github.com> * β¬οΈ Use new process release * π§ Get EUDEMO running (#2752) * π§ WIP get eudemo working * π¨ Neater offset fix * π¨ Safe xxw * π 'stabilise' coil supports optimisation * PROCESS inputs tweaks for EU-DEMO (#2771) * safety factor to 3.5 * CS fatigue inputs to SS316LN * CS fatigue inputs to SS316LN * working set? * π¨ Jc converges --------- Co-authored-by: james <james.cook1@ukaea.uk> * π¨ + marker for coils with no current * π¨ Make xxw easier to decipher --------- Co-authored-by: CoronelBuendia <matti.coleman@gmail.com> --------- Co-authored-by: CoronelBuendia <matti.coleman@gmail.com>
- Loading branch information