Status

Changes

Loading builds...

Changes

#32 (Jul 16, 2026, 11:52:00 PM)

CHG: Support string cell arrays for colorbar ticks. Additionally, generate numeric tick labels on a single line.

inwoopark0415 at 11:44 PM 7/9/26

CHG: adding HydrologyMoulinInputEnum as control and a couple of estetic changes

Mathieu Morlighem at 3:15 AM 7/10/26

BUG: when copying a hook, we should NOT copy the pointers because other datasets are already being copied and we don't want to point to obects of the original dataset

Mathieu Morlighem at 3:31 AM 7/10/26

BUG: now loads need to be reconfigured after being copied

Mathieu Morlighem at 4:01 AM 7/10/26

CHG: Fix healing term in damage evolution with Pralong.

inwoopark0415 at 5:38 PM 7/11/26

CHG: Fix healing term in damage evolution with Pralong.

inwoo0415 at 5:47 PM 7/11/26

CHG: Update the archive for test275 to reflect the changes to the Pralong damage law.

inwoopark0415 at 6:51 PM 7/11/26

CHG: added comment sign so that output name display when typing help

mathieu.morlighem at 11:55 AM 7/13/26

BUG: forgot to delete materials for 3D meshes

Mathieu Morlighem at 8:24 AM 7/15/26

CHG: Improved sealevelchange core's tracking of mass changes when significant grounding/coast line migration is involved

lambert.caron at 9:27 AM 7/16/26

CHG: Improved sealevelchange core's tracking of mass changes when significant grounding/coast line migration is involved

lambert.caron at 9:29 AM 7/16/26

CHG: Improved sealevelchange core's tracking of mass changes when significant grounding/coast line migration is involved

lambert.caron at 11:10 AM 7/16/26

CHG: Improved sealevelchange core's tracking of mass changes when significant grounding/coast line migration is involved

lambert.caron at 11:42 AM 7/16/26

[BUG]: bamg seems to be OS-dependent when it comes to the mesh for test2091, changed the test to check a profile on required vertices

lambert.caron at 1:25 PM 7/16/26

[BUG]: switching to pre-gen mesh for test2091, 2092 for consistency across OS

lambert.caron at 2:03 PM 7/16/26

#31 (Jul 9, 2026, 11:52:00 PM)

CHG: no need to pull C every iteration

Mathieu Morlighem at 3:41 AM 7/3/26

Simplified ISSM and PETSc configs for Windows

34657535+TidbitSoftware at 11:14 AM 7/3/26

Lowering C standard for GCC when compiling Triangle on Windows

34657535+TidbitSoftware at 3:04 PM 7/3/26

Reducing list of manually-supplied library link flags now that we are using newer pacman-supplied copy of MS-MPI

34657535+TidbitSoftware at 7:45 PM 7/3/26

Need quadmath library

34657535+TidbitSoftware at 10:51 PM 7/3/26

BUG: tiny fix, missing the 3 parameter case

f007fk5 at 2:26 PM 7/4/26

CHG: cleanup: removing Radar cost function and Enums, and removing triangle.h from toolkits

mathieu.morlighem at 9:07 AM 7/6/26

CHG: no more radar cost function

mathieu.morlighem at 9:14 AM 7/6/26

CHG: removing header file

mathieu.morlighem at 2:33 AM 7/9/26

CHG: added example

Mathieu Morlighem at 2:34 AM 7/9/26

CHG: relaxed tolerances

mathieu.morlighem at 2:52 AM 7/9/26

CHG: Remove user's specific.

inwoo0415 at 6:23 PM 7/9/26

CHG: Update default parameters in ismip7 ocean forcings.

inwoo0415 at 6:24 PM 7/9/26

#30 (Jul 2, 2026, 11:52:00 PM)

CHG: removed lines not used

helene.l.seroussi at 3:25 PM 6/29/26

NEW: added module for ocean connectivity and ice shelf melt applied only to areas connected to the ocean

helene.l.seroussi at 4:26 AM 6/30/26

FIX: need new nomelt_under_lakes for vertical analysis

helene.l.seroussi at 5:07 AM 6/30/26

FIX: forgot bracket for the check because of nomelt_under_lakes

helene.l.seroussi at 5:20 AM 6/30/26

FIX: fixed new values check in nomeltunderlake for python and js

helene.l.seroussi at 5:54 AM 6/30/26

CHG: interpISMIP7AntarcticaSMB.m - Fix treating nan values in SMB. Also enable to support 'historical' periods.

inwoo0415 at 6:29 AM 6/30/26

BUG: Resolve memory leakage problem in Ismip7FloatingiceMeltingRate.

inwoo0415 at 6:29 AM 6/30/26

Updates for Windows builds after upgrading system copy of Python

34657535+TidbitSoftware at 1:45 PM 7/1/26

CHG: added better error message so that we can remove this function altogether

mathieu.morlighem at 3:41 PM 7/1/26

CHG: minor cosmetics

mathieu.morlighem at 3:41 PM 7/1/26

CHG: minor cosmetics

mathieu.morlighem at 3:41 PM 7/1/26

NEW: adding new way to read and broadcast data from binary file. Until now rank 0 reads and broadcasts entire fields but we sometimes run out of memory for very large fields (e.g., SMB). Instead, rank 0 now creates tailored arrays for each partition and that's what's being send. It does not work yet, this is an intermediary commit for debugging, so this behavior is not yet activated. This will be a long summer project...

mathieu.morlighem at 3:45 PM 7/1/26

NEW: new approach now working and debugged, but not turned on by default

Mathieu Morlighem at 4:36 AM 7/2/26

BUG: only delete rank0_vert_ids if it is not NULL

Mathieu Morlighem at 11:17 AM 7/2/26

#29 (Jun 25, 2026, 11:52:00 PM)

Update basalforcingsismip6.m to fix wrong default value for 'islocal'

noreply at 7:02 AM 6/19/26

CHG: trying to fix nudging class (needs to pull out fields from md.inversion

Mathieu Morlighem at 7:55 AM 6/19/26

CHG: mech/mechanicalproperties.m updated to ensure that power (n) cannot be of length 1.

benjamingetraer at 10:33 AM 6/22/26

CHG: added RMSEs as J in results

mathieu.morlighem at 10:34 AM 6/23/26

CHG: minor cosmetics for valgrind

Mathieu Morlighem at 10:46 AM 6/23/26

CHG: no need to prescribe 'benchmark','all' when prescriving ids>1000

mathieu.morlighem at 1:16 PM 6/24/26

#28 (Jun 18, 2026, 11:52:00 PM)

CHG: added RMSE's to results as J

mathieu.morlighem at 3:50 PM 6/12/26

CHG: interpISMIP7AntarcticaOcn.m - Fix to check time boundary and update parameters for ISMP7.

inwoo0415 at 4:33 AM 6/13/26

CHG: interpISMIP7AntarcticaSMB.m - Use anomaly values for SMB and reference SMB from RACMO24p1 dataset.

inwoo0415 at 11:12 PM 6/13/26

BUG: Fix bug in interpISMIP7AntarcticaOcn.m.

inwoo0415 at 12:38 AM 6/14/26

CHG: new way of computing RMSEs for nudging

mathieu.morlighem at 10:58 AM 6/18/26

CHG: do we REALLY need VxObsEnum? trying to use InversionVxObs instead, it makes more sense to reuse existing Enums if possible

mathieu.morlighem at 10:58 AM 6/18/26

#27 (Jun 11, 2026, 11:52:00 PM)

CHG: clean up comment in GEMB accumulation

95945328+NJSchlegel at 6:24 PM 5/21/26

CHG: added melt to nudging scheme

Mathieu Morlighem at 5:01 AM 6/5/26

CHG: fixing melt nudging

mathieu.morlighem at 5:15 PM 6/5/26

CHG: fixed dimension problem in nudging scheme

mathieu.morlighem at 4:06 PM 6/6/26

NEW: added GetVectorFromInputsx and UpdateInputFromVector with Lid so that we can do this locally

mathieu.morlighem at 2:29 PM 6/7/26

CHG: minor estaetics

mathieu.morlighem at 2:33 PM 6/7/26

CHG: GEMB read mapping neighbors as double array

95945328+NJSchlegel at 9:17 AM 6/8/26

CHG: added warning when control exceeds parameter_min/max and gradient is 0

mathieu.morlighem at 5:40 PM 6/8/26

CHG: there was a bug when using git difftool (error message complaining about path), this is now fixed

mathieu.morlighem at 3:44 PM 6/9/26

NEW: you can now customize where input files are created using md.settings.stagingpath

mathieu.morlighem at 5:08 AM 6/10/26

mathieu.morlighem at 5:20 AM 6/10/26

Reducing lib flag list so that compiler can find the correct versions on its own.

34657535+TidbitSoftware at 1:10 PM 6/10/26

Reverting lib flags change and rebuilding ext pkgs

34657535+TidbitSoftware at 1:46 PM 6/10/26

NEW: added master flag to not save results in a transient run, needed for nudging

mathieu.morlighem at 6:42 PM 6/11/26

CHG: only save last step

mathieu.morlighem at 6:57 PM 6/11/26

CHG: BasalforcingsPerturbationMeltingRate output is now in m/yr

mathieu.morlighem at 11:25 PM 6/11/26

NEW: added NR test for nudging

mathieu.morlighem at 11:31 PM 6/11/26

#26 (Jun 4, 2026, 11:52:00 PM)

issmversion.py: Fix so docstring is correct and prints version

ken.mankoff at 4:06 AM 6/1/26

export_netCDF.py: Add overwrite keyword

ken.mankoff at 6:44 AM 6/1/26

CHG: shielding seminc from ADOLC

Mathieu Morlighem at 10:11 AM 6/1/26

CHG: moving constant to ifdef

Mathieu Morlighem at 11:28 AM 6/1/26

remove duplicate gadi file from cluster dir

justinkinjun.hew at 4:35 PM 6/1/26

NEW: added cost function 701, divergence^2, to improve model initialization

Mathieu Morlighem at 9:56 AM 6/2/26

CHG: when running valgrind make sure that outlog and errlog are in the right directory

Mathieu Morlighem at 10:04 AM 6/2/26

CHG: need to add a few libraries to LD_LIBRARY_PATH in queue script

Mathieu Morlighem at 10:06 AM 6/2/26

CHG: completing first version of nudging code

mathieu.morlighem at 1:22 PM 6/2/26

CHG: outputing friction automatically

mathieu.morlighem at 1:32 PM 6/2/26

BUG: could not run FS with complete adjoint

mathieu.morlighem at 9:49 PM 6/2/26

CHG: keep code in SI (s) only convert to m/yr in RMSE

mathieu.morlighem at 7:25 PM 6/3/26

CHG: added more options for nudging

mathieu.morlighem at 8:02 PM 6/3/26

#25 (Jun 1, 2026, 1:36:10 AM)

CHG: streamlining cluster classes

mathieu.morlighem at 4:55 AM 5/22/26

CHG: port number should be an integer, not a string

mathieu.morlighem at 4:57 AM 5/22/26

BUG: fixing CoDiPack by converting PDouble to Double

Mathieu Morlighem at 5:52 AM 5/22/26

CHG: removing LaunchQueueJobIceOcean, we can call the usual LaunchQueueJob

mathieu.morlighem at 6:44 AM 5/22/26

Testing if use of new generic class will suffice for binaries

34657535+TidbitSoftware at 11:08 AM 5/22/26

Removed static version of generic class; adjusted tolerances

34657535+TidbitSoftware at 4:37 PM 5/22/26

Forgot to update Python binaries after removing static generic cluster class

34657535+TidbitSoftware at 4:59 PM 5/22/26

Also needed to clean up import and check on static generic class in waitonlock

34657535+TidbitSoftware at 5:01 PM 5/22/26

Adjusted tolerances

34657535+TidbitSoftware at 6:35 PM 5/22/26

CHG: streamlining QueueScript so that new clusters don't have a ton of code to write

mathieu.morlighem at 4:06 AM 5/24/26

BUG: restored classes/clusters/cluster_defaults.py

mathieu.morlighem at 4:12 AM 5/24/26

CHG: fixing dakota

mathieu.morlighem at 7:30 AM 5/24/26

CHG: gixing dakota version in python

Mathieu Morlighem at 8:18 AM 5/24/26

CHG: fixing dakota

mathieu.morlighem at 12:28 PM 5/24/26

CHG: crop anything outside the data domain when interpolating dHdt

enigne at 6:28 PM 5/24/26

CHG:add options to data loaders to load from specific nc files

enigne at 10:25 PM 5/24/26

CHG: moved SEMIC from externalpackages to C++ in src/c/shared

mathieu.morlighem at 4:42 AM 5/26/26

CHG: SEMIC is not a required external package anymore

mathieu.morlighem at 7:37 AM 5/26/26

CHG: missing header file to get machine eps

mathieu.morlighem at 7:50 AM 5/26/26

CHG: replaced double by IssmDouble for CoDiPack

mathieu.morlighem at 8:06 AM 5/26/26

BUG: problem recasting IssmDouble to int

Mathieu Morlighem at 9:20 AM 5/26/26

CHG: updating now that semic and m1qn3 are natively in ISSM

Mathieu Morlighem at 10:00 AM 5/26/26

CHG: Now, a unit for TotalHydrologyFlux at grounding line is Gt yr-1.

inwoo0415 at 10:43 PM 5/27/26

BUG: Fix typos....

inwoo0415 at 12:24 AM 5/28/26

CHG: processdata.py - Fix treating element dataset.

inwoo0415 at 1:16 AM 5/28/26

CHG: cmaptools.py - Add functionality to handle LinearSegmentedColormap and string.

inwoo0415 at 10:31 PM 5/31/26

CHG: Fix typos.

inwoo0415 at 10:55 PM 5/31/26

BUG: Fix typos.

inwoo0415 at 10:56 PM 5/31/26

#24 (May 21, 2026, 11:52:00 PM)

CHG: eliminating m1qn3 from Github's CI

mathieu.morlighem at 9:53 AM 5/15/26

CHG: now call MAR-IA once per processor instead of per element

55120383+yinmin-liu at 10:51 AM 5/15/26

NEW: added clusters/unity.m new unity cluster

mathieu.morlighem at 4:26 PM 5/17/26

NEW: consolidated read_netCDF/write_netCDF should now be more robust:

Mathieu Morlighem at 4:30 PM 5/17/26

CHG: removing extraneous printf

Mathieu Morlighem at 4:33 PM 5/17/26

CHG: removing extraneous printf

Mathieu Morlighem at 4:33 PM 5/17/26

CHG: streamlining generic queueing script

Mathieu Morlighem at 4:34 PM 5/17/26

CHG: moving old code to OLD to avoid confusion

Mathieu Morlighem at 4:44 PM 5/17/26

BUG: divergence.m - 2-D mesh does not require z-coordinate.

inwoo0415 at 7:14 AM 5/18/26

CHG: model is now marshalled in execution, need to take that into account when scp'ing

mansakrishna23 at 1:43 PM 5/18/26

CHG: fixed UploadQueueScript now that all the files are prepared in execution

Mathieu Morlighem at 5:15 PM 5/18/26

CHG: these classes do not have a port property, replace with 0

Mathieu Morlighem at 5:20 PM 5/18/26

BUG: divergence.m - 2-D mesh does not require z-coordinate.

inwoo0415 at 6:21 PM 5/18/26

CHG: model is now marshalled in execution, need to take that into account when scp'ing

inwoo0415 at 6:22 PM 5/18/26

CHG: fixed UploadQueueScript now that all the files are prepared in execution

inwoo0415 at 6:22 PM 5/18/26

CHG: plot_gridded.py - Clean up code and fix bug in getcolormap.

inwoo0415 at 6:30 PM 5/18/26

CHG: get status from copyfile so that MATLAB does not error out, and display a warning instead

mathieu.morlighem at 9:36 AM 5/19/26

CHG: Synchronize Matlab > Python. Display a warning message if we cannot copy the file.

inwoo0415 at 6:39 PM 5/19/26

CHG: Fix minor.

inwoo0415 at 6:41 PM 5/19/26

Adjusted tolerances

34657535+TidbitSoftware at 7:53 PM 5/19/26

CHG: Fix minor.

inwoo0415 at 11:01 PM 5/19/26

CHG: a lot of code in clusters is repetitive, move Upload/Launch/Download in cluster_defaults to simplify the implementation of new cluster classes

mathieu.morlighem at 5:47 AM 5/20/26

CHG: streamlining all cluster classes with new defaults

mathieu.morlighem at 5:02 AM 5/21/26

NEW: adding inversionnudging for Tim's work

mathieu.morlighem at 9:00 AM 5/21/26

CHG: no need to get min/max parameters when creating new elements

Mathieu Morlighem at 9:19 AM 5/21/26

CHG: we do not need InputUpdateFromIoModel anymore

Mathieu Morlighem at 9:27 AM 5/21/26

CHG: ptol is too high in GEMB

95945328+NJSchlegel at 4:32 PM 5/21/26

CHG: GEMB test removal of P tol in accumulation

95945328+NJSchlegel at 6:02 PM 5/21/26

CHG: working on nudging method

Mathieu Morlighem at 6:13 PM 5/21/26

#23 (May 14, 2026, 11:52:00 PM)

NEW: m1qn3 now returns the BEST X/G that it found

mathieu.morlighem at 5:50 AM 5/8/26

NEW: added support for 3.25

mathieu.morlighem at 5:50 AM 5/8/26

BUG: Fix bugs in interpISMIP7AntarcticaSMB.m.

inwoo0415 at 6:43 PM 5/10/26

CHG: making Bamg more robust to produce the same mesh across platform. Right now meshes are slightly different on mac and linux. This commit should fix it

mathieu.morlighem at 4:58 AM 5/11/26

CHG: updated Archives now that bamg is more portable

Mathieu Morlighem at 5:25 AM 5/11/26

CHG: adding more tests to Mac M now that bamg is more portable

Mathieu Morlighem at 5:28 AM 5/11/26

CHG: updated archives now that bamg is more robust

Mathieu Morlighem at 5:51 AM 5/11/26

CHG: still erroring out on bamg adaptive mesh

Mathieu Morlighem at 9:16 AM 5/11/26

CHG: excluding test703 again because it uses bamg

Mathieu Morlighem at 9:38 AM 5/11/26

CHG: offset message to be consistent with other notifications

mathieu.morlighem at 9:40 AM 5/11/26

NEW: now create files directly in execution/ instead of current directory to avoid clutter and overwrites

mathieu.morlighem at 1:03 PM 5/11/26

BUG: issmscpin.py – Fix missing bracketstyle argument in the function call. Also fix the missing path variable when constructing the SCP command.

inwoo0415 at 7:23 PM 5/11/26

BUG: issmscpin.py – Fix missing bracketstyle argument in the function call. Also fix the missing path variable when constructing the SCP command.

inwoo0415 at 8:35 PM 5/11/26

BUG: handle case when executionpath is not /Users/mmorligh/Desktop/ISSM/execution

mathieu.morlighem at 4:56 AM 5/12/26

CHG: better way to handle local vs remote execution

mathieu.morlighem at 7:02 AM 5/12/26

CHG: trying to fix windows with the new format of execution/input files

mathieu.morlighem at 7:27 AM 5/12/26

CHG: fixing windows

mathieu.morlighem at 10:20 AM 5/12/26

m1qn3 is now integrated

34657535+TidbitSoftware at 1:31 PM 5/12/26

Removing m1qn3 fence; formatting

34657535+TidbitSoftware at 4:03 PM 5/12/26

m1qn3 is now integrated

34657535+TidbitSoftware at 5:00 PM 5/12/26

CHG: in some cases the new m1qn3 would fail with exit code 7, this should now be fixed

mathieu.morlighem at 4:45 AM 5/13/26

CHG: minor update

mathieu.morlighem at 4:46 AM 5/13/26

CHG: trying to fix dakota now that files are explicitly in execution

mathieu.morlighem at 5:41 AM 5/13/26

NEW:preparing a GNN based stressbalance emulator

55120383+yinmin-liu at 9:39 AM 5/13/26

frontera, andes settings

mansakrishna23 at 12:59 PM 5/13/26

CHG: use Dakota's output manager to figure out where to read and write files

mathieu.morlighem at 4:36 AM 5/14/26

CHG: added check to warn users when execution dir needs to be cleaned up

mathieu.morlighem at 4:56 AM 5/14/26

BUG: adding default pertubration melt for spatial linear

mansakrishna23 at 12:39 PM 5/14/26

CHG: make sure to save outlog and errlog in execution directory if launched from another directory (no more cd)

mansakrishna23 at 12:40 PM 5/14/26

Delete src/m/andes_settings.m

noreply at 1:24 PM 5/14/26

Delete src/m/frontera_settings.m

noreply at 1:24 PM 5/14/26

Restore MATLAB export_netCDF.m

noreply at 5:16 PM 5/14/26