Open Hub
Open Hub

 Settings |  Report Duplicate

1
I Use This!
Very Low Activity

Commits : Listings

Analyzed 21 days ago. based on code collected 22 days ago.
Jan 29, 2024 — Jan 29, 2025
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
these changes affect IceModel (and all derived): * ColumnSystemCtx base class now holds column indices and can reset system * some updated/corrected comments in grid.cc, doc/browser_pages.txt
bueler
as Ed Bueler
More... about 15 years ago
* Fixed the "showEnv" make target (it was not printing PISM_CONFIG). * Added a little bit of code to allow flushing time-series by sending USR2 to a running PISM process (or processes). * Fixed the history string handling. * Fixed IceModelVec3::getPlaneStarZ() (another bug caused by using uninitialized memory). * Improved IceModelVec3::has_nan().
Constantine Khroulev
More... about 15 years ago
* removal of BasalTypeSIA (a type in materials.cc|hh) and BasalTypeSIA *basalSIA from IceModel
bueler
as Ed Bueler
More... about 15 years ago
* traded in getEffectivePressureOnTill() for getBasalWaterPressure(); the former is easy to calculate from the later if needed
bueler
as Ed Bueler
More... about 15 years ago
THIS COMMIT MESSAGE SHOWS MOST r927:928 CHANGES AS WELL AS r928:929 (sorry ... bad command in prev commit ...)
bueler
as Ed Bueler
More... about 15 years ago
Suggested move of initBasalTillModel() from misc_setup() to model_state_setup(). This seems to work, but moving bed deformation setup seems not to work; why does it seg fault? See preprocessor directives in src/base/iMinit.cc.
bueler
as Ed Bueler
More... about 15 years ago
Added executable permission to nc2mat.py.
bueler
as Ed Bueler
More... about 15 years ago
Minor updates to EISMINT-Greenland outputs now that year 0 diagnostics are valid.
bueler
as Ed Bueler
More... about 15 years ago
Added initializations for two more IceModel parameters.
Constantine Khroulev
More... about 15 years ago
* Fixed a bug in IceModel::SigmaSIAToRegular() (look for "<=" instead of "<" around iMsia.cc:516). Yes, this one-character change seems to fix pismv -test K, the "pisms; pismr -i simp_exper.nc" sequence and software test 12 (penth). * Modified util/nccmp.py to handle files with NANs more gracefully. * Improved the "diagnostic" step and model re-initialization in IceModel::run(). * Added PISMVars::check_for_nan() and IceModelVec::has_nan() (for debugging). * Slightly improved command-line option processing (options -ice_type, -z_spacing, -zb_spacing).
Constantine Khroulev
More... about 15 years ago
- modified buil/Makefile to make it work with petsc-dev - renamed env variable CONFIG -> PISM_CONFIG. It invokes additional OS-specific options into Makefile, e.g. used for OS X. Renaming avoids possible clashes with other software (such as petsc-dev). Modified installation manual accordingly.
Andy Aschwanden
More... about 15 years ago
- fixed (hopefully) bug in python scripts which create a new netCDF file. Using netCDF4 module caused creating a netCDF4 file by default. This was changed to NETCDF3_CLASSIC instead.
Andy Aschwanden
More... about 15 years ago
Fixed summaryPrintLine() to deal correctly with -no_temp runs.
bueler
as Ed Bueler
More... about 15 years ago
* Florian is distinctly right about bug #13974.
bueler
as Ed Bueler
More... about 15 years ago
* Created IceModel::step(...); moved almost all the contents of the time-stepping loop there. * Modified the initialization sequence so that all the code initializing model state variables is in (or is called from) IceModel::model_state_setup(). Still need to check derived classes of IceModel. * Modified IceModel::run() so that PISM 1) does one time-step 2) re-sets time and re-initializes model state variables 3) goes on to the time-stepping loop. This way a -y 0 run is (should be) equivalent to a "diagnostic" run. Also, this extra time-step makes it possible to compute rates-of-change at the beginning of the run. * Fixed the util/fill_missing.py file permission issue (an output file was only readable by its owner).
Constantine Khroulev
More... about 15 years ago
* Moved "config" option database out of IceModel and into main(). This guarantees that it (the config database) is initialized *before* it is used. * Changed build/Makefile to import PETSc make variables (and not rules). Will revert to r917 if this does not work. * Removed the code initializing mapping parameters from IceModel::setDefaults(): there is no sensible default I know of.
Constantine Khroulev
More... about 15 years ago
- added -init_from_temp_liqfrac option for debugging Storglaciaren melt rate bug. Option is similar to -init_from_temp, but calculates enthalpy from temperature and liquid water fraction found in the input file
Andy Aschwanden
More... about 15 years ago
Made IceEnthalpyModel method of computing effective pressure on till, potentially-using basal melt rate, into standard IceModel version. Renamed a method to updateYieldStressUsingBasalWater(). Fixed IceEnthalpyModel::enthalpyAndDrainageStep() to correctly treat case where k0+ks==0. Re-ordered initialization procedures in iceEnthalpyModel.cc to follow actual initialization order. (None of these seem to help with bugs.)
bueler
as Ed Bueler
More... about 15 years ago
Fleshed out init sequence flow chart.
bueler
as Ed Bueler
More... about 15 years ago
In attempt to track down penth bug, I have improved nccmp.py (I think). And I have added -init_from_temp option to IceEnthalpyModel.
bueler
as Ed Bueler
More... about 15 years ago
Now that Constantine has made the EISMINT-Ross example work again, I have done some additional minor maintenance.
bueler
as Ed Bueler
More... about 15 years ago
Minor maintenance: prepend to history string in fill_missing.py; fixed messages in iMtimeseries.cc
bueler
as Ed Bueler
More... about 15 years ago
Test I was broken in 908:909 changes, I think. Turns out Test I was using uninitialized vubarSSA,vvbarSSA. Now have_ssa_velocities is explicitly set to false by IceExactSSAModel.
bueler
as Ed Bueler
More... about 15 years ago
Only minor fixes to stdout messages, documentation, and the long name of acab.
bueler
as Ed Bueler
More... about 15 years ago
* Fixed EISMINT-Ross and rossplot.py (max computed speed reported: 1378.185 (m/a), average relative error in vector vel = 0.07279, chi^2 = 3684.510) * IceModelVecs (-2 and -3Bedrock) are initialized to NAN at allocation time: this way we can easily see it a field was not initialized at the model setup stage. Added _FillValue as a default NetCDF attribute. Modified iceEISModel.cc and iceCompModel.cc to initialize tillphi explicitly. * Renamed software test #1 to "pismr (SIA only) exact restartability".
Constantine Khroulev
More... about 15 years ago
Looked at EISMINT-Ross example; definitely broken from deprecated use of 'fundamental transpose', which can now be eliminated. TO DO, that is; I have left some FIXMEs to help on some issues, too.
bueler
as Ed Bueler
More... about 15 years ago
Attempting to make sure that basal melt rate (bmelt=vbasalMeltRate) is properly-initialized. Also trivial stdout change for penth.
bueler
as Ed Bueler
More... about 15 years ago
Doc improvements only: editing of BOMBPROOF notes and revision of -verbose option docs.
bueler
as Ed Bueler
More... about 15 years ago
Fixes to grid parameter reporting and verbose option.
bueler
as Ed Bueler
More... about 15 years ago
Cleaned up and simplified the standard out flags and summary. Now that we have diagnotic viewers, less crap should be dumped to stdout.
bueler
as Ed Bueler
More... about 15 years ago