Skip to content
[Jenkins]Jenkins
log in
Success

Changes

Summary

  1. I have turned the nu_vert_res_dep family of variables into a derived (details)
  2. Add 3 new features to dashboard: (details)
Commit 1b4ca12287908273f189d8588e7a39755cd4fb6d by bmg2
I have turned the nu_vert_res_dep family of variables into a derived
type and that variable is now being passed throughout the CLUBB code,
rather than accessing variables through a USE statement.

Everything is 100% bit-for-bit!
The file was modified src/CLUBB_core/advance_xm_wpxp_module.F90 (diff)
The file was modified src/advance_microphys_module.F90 (diff)
The file was modified src/CLUBB_core/parameters_tunable.F90 (diff)
The file was modified src/CLUBB_core/advance_clubb_core_module.F90 (diff)
The file was modified src/G_unit_test_types/spurious_source_test.F90 (diff)
The file was modified src/CLUBB_core/advance_windm_edsclrm_module.F90 (diff)
The file was modified src/CLUBB_core/advance_xp2_xpyp_module.F90 (diff)
The file was modified src/CLUBB_core/advance_wp2_wp3_module.F90 (diff)
The file was modified src/clubb_driver.F90 (diff)
The file was modified src/CLUBB_core/clubb_api_module.F90 (diff)
The file was modified src/CLUBB_core/diffusion.F90 (diff)
Commit ae681935179f62ffb588fe2aab7c95780034f153 by Vince Larson
Add 3 new features to dashboard:
1) Add Huber (confusingly called "Ransac") regressor to find
a fit that de-weights outliers.  The goal is to identify tuning trade-offs.
2) Add ElasticNet regressor.  The goal here is to obtain
a good fit that doesn't perturb the parameters too much,
and that leaves some parameter values unchanged if possible.
3) Add a routine to create PCA biplots.  The goal here
is to identify outliers (i.e., tuning trade-offs) analytically.

For #910.
The file was modified utilities/sens_matrix/sens_matrix_dashboard.py (diff)
The file was modified utilities/sens_matrix/test_analyzeSensMatrix.py (diff)
The file was modified utilities/sens_matrix/analyze_sensitivity_matrix.py (diff)