Started by an SCM change12912407142b14155687d0d80b174796965d2793b1212407142b14155687d0d80b174796965d2793b12origin/master12407142b14155687d0d80b174796965d2793b1212407142b14155687d0d80b174796965d2793b12origin/master113b19b67713baf175f99baa0cb0bdde6e8c1fbc588b19b67713baf175f99baa0cb0bdde6e8c1fbc588origin/grid_fixesb19b67713baf175f99baa0cb0bdde6e8c1fbc588b19b67713baf175f99baa0cb0bdde6e8c1fbc588origin/grid_fixes108b928767c7238139d599dc55420d2a0956161726ab928767c7238139d599dc55420d2a0956161726aorigin/feeding_gr_to_hostb928767c7238139d599dc55420d2a0956161726ab928767c7238139d599dc55420d2a0956161726aorigin/feeding_gr_to_host1586074d7b38f6bd68b4344ab19c9e3ddf8de415e956074d7b38f6bd68b4344ab19c9e3ddf8de415e95refs/remotes/origin/master6074d7b38f6bd68b4344ab19c9e3ddf8de415e956074d7b38f6bd68b4344ab19c9e3ddf8de415e95refs/remotes/origin/master6074d7b38f6bd68b4344ab19c9e3ddf8de415e956074d7b38f6bd68b4344ab19c9e3ddf8de415e95refs/remotes/origin/masterhttps://github.com/larson-group/clubb.git1589c672f54d3fb0c316475258609ebd1c70b2d3bca9c672f54d3fb0c316475258609ebd1c70b2d3bcarefs/remotes/origin/master9c672f54d3fb0c316475258609ebd1c70b2d3bca9c672f54d3fb0c316475258609ebd1c70b2d3bcarefs/remotes/origin/master9c672f54d3fb0c316475258609ebd1c70b2d3bca9c672f54d3fb0c316475258609ebd1c70b2d3bcarefs/remotes/origin/masterhttps://github.com/larson-group/sys_adminfalse#1585033621912456clubb_fire_tuning_gfortran_test #158158false1588600SUCCESS1630139767693http://carson.math.uwm.edu/jenkins/job/clubb_fire_tuning_gfortran_test/158/utilities/sens_matrix/test_analyzeSensMatrix.pyutilities/sens_matrix/analyze_sensitivity_matrix.pyutilities/sens_matrix/sens_matrix_dashboard.py6074d7b38f6bd68b4344ab19c9e3ddf8de415e951630082416000http://carson.math.uwm.edu/jenkins/user/vlarsonVince Larsonvlarson@uwm.eduChange normalization of sensitivity matrix. Before it was normalized by (default-observed) values of metrics. Now it is normalized by 1/obs. This leaves the sensitivity matrix free from dependence on the bias and puts the bias entirely on the right-hand side of the matrix equation. Also, the parameter values are no longer normalized by max(def,sens) values. Instead, they are normalized by the values used in the default simulation, and a sensitivity simulation value is used only if the default-simulation value is zero. Add plots of left-singular vectors and right-singular vectors. Annotate these plots with singular values. One of the unit tests, test_3x2_C8transformed, is broken. For #910. 2021-08-27 11:40:16 -05006074d7b38f6bd68b4344ab19c9e3ddf8de415e95Change normalization of sensitivity matrix.editutilities/sens_matrix/test_analyzeSensMatrix.pyeditutilities/sens_matrix/analyze_sensitivity_matrix.pyeditutilities/sens_matrix/sens_matrix_dashboard.pygithttp://carson.math.uwm.edu/jenkins/user/vlarsonVince Larsonfalse159http://carson.math.uwm.edu/jenkins/job/clubb_fire_tuning_gfortran_test/159/157http://carson.math.uwm.edu/jenkins/job/clubb_fire_tuning_gfortran_test/157/