Code

Add ML coupling#14

Code

Open

jatkinson1000 wants to merge 22 commits intodatawave_mlDataWaveProject/CAM:datawave_mlfrom couplingDataWaveProject/CAM:couplingCopy head branch name to clipboard

+519-16

Commits

Commits on Aug 19, 2024

• Add variable for the convect_dp_gw net filepath to be read from namelist.

  jatkinson1000

  committed
  fd6e669

  Copy full SHA for fd6e669

• WIP Adapt gw_drag.F90 to import ftorch and read in a net from file. TODO: Move reading net to init routine.

  jatkinson1000

  committed
  bed301e

  Copy full SHA for bed301e

• Move reading of NN to initialisation routine for gw_drag.

  jatkinson1000

  committed
  7968fa9

  Copy full SHA for 7968fa9

• Add a gw_final() subroutine to destroy NN instance.

  jatkinson1000

  committed
  b3ac3e9

  Copy full SHA for b3ac3e9

• Remove spurious test code.

  jatkinson1000

  committed
  16b53ac

  Copy full SHA for 16b53ac

• Update source to use latest version of FTorch following API Changes.

  jatkinson1000

  committed
  dfd451d

  Copy full SHA for dfd451d

• WIP add gw_ml file with starting point for a routine to run the ML parameterisation via FTorch.

  Show description for 0d334db
  jatkinson1000

  committed
  0d334db

  Copy full SHA for 0d334db

• Add normalisation file to namelist variables, read, and distribute.

  jatkinson1000

  committed
  01950bc

  Copy full SHA for 01950bc

Commits on Oct 25, 2024

• Create gw_ml initialisation routine to read in net and normalisation variables from file.

  jatkinson1000

  committed
  603c78a

  Copy full SHA for 603c78a

• Add finalisation routine to gw_ml to destroy the net.

  jatkinson1000

  committed
  72a50a9

  Copy full SHA for 72a50a9

• Add normalisation routine for inputs to net.

  jatkinson1000

  committed
  a9f95ee

  Copy full SHA for a9f95ee

• Add denormalisation routine for the gw_convect_dp net.

  jatkinson1000

  committed
  19fa91f

  Copy full SHA for 19fa91f

• BugFix: Remove redundant net variable

  jatkinson1000

  committed
  7bf43e1

  Copy full SHA for 7bf43e1

• Couple gw_drag to the ML scheme

  jatkinson1000

  committed
  c01db7e

  Copy full SHA for c01db7e

• Bugfix: Update ML switches to be logicals when calling gw_drag init and finalize routines.

  jatkinson1000

  committed
  921a189

  Copy full SHA for 921a189

• [DEBUG: Run net with ones as the input to compare to same operation using pure python (Same result obtained).]

  jatkinson1000

  committed
  aaecc4a

  Copy full SHA for aaecc4a

• Revert changes to debug commit using ones as net input.

  jatkinson1000

  committed
  5990616

  Copy full SHA for 5990616

• Add output for U and V tendencies from GW from Net and Beres schemes.

  jatkinson1000

  committed
  83cb8c9

  Copy full SHA for 83cb8c9

• Set tendencies not predicted by the net to 0.0 for the ml scheme rather than running the physics scheme beforehand.

  jatkinson1000

  committed
  03bd7f5

  Copy full SHA for 03bd7f5

• Add iulog to gw_ml to write outputs to diagnostics in init scheme.

  jatkinson1000

  committed
  bf2a8bb

  Copy full SHA for bf2a8bb

• Add code to produce output required by the testcase.

  jatkinson1000

  committed
  218242f

  Copy full SHA for 218242f

• Update rhoi output to be ncol rather than pcol.

  jatkinson1000

  committed
  a2d881c

  Copy full SHA for a2d881c