check surface jacobian after recent PR #149

[BenjaminTJohnson]

  4 - test_Unit_Aerosol_Bypass (Failed)
  5 - test_Unit_Aerosol_Bypass_TL (Failed)
  6 - test_Unit_Aerosol_Bypass_adjoint (Failed)
  7 - test_Unit_Aerosol_Bypass_k_matrix (Failed)

These fail in develop after recent PR #149 when using ctest results from prior to PR.

7:      Comparing calculated results with saved ones...
7:  test_Aerosol_Bypass_k_matrix(FAILURE) : Surface_K Jacobians are different!

@chengdang , can you test and verify? run ctests on commit prior to PR 149, then run ctests on develop without removing results.

[chengdang]

Hello @BenjaminTJohnson, thank you for your tests!
Somehow I could not reproduce the error. All ctest passed for me comparing develop branch versus the PR150 commit ( b305b26). Was this the commit you compared against to? Thanks.

[BenjaminTJohnson]

you can run the ctests in develop, keep the results in the build directory, check out another check out a different commit (e.g., 262599d) , compile and re-run the ctest(s).

[chengdang]

That's what I did, one commit prior to PR #149 was the PR 150 commit (b305b26)

I can try again with a much earlier commit 262599d

[chengdang]

Here are the ctests results, comparing to the result calculated using develop branch:

With commit 262599d, 39ed197

 99% tests passed, 2 tests failed out of 162

Label Time Summary:
CRTM_Tests    = 184.58 sec*proc (162 tests)

Total Test time (real) = 188.78 sec

The following tests FAILED:
	  4 - test_Unit_Aerosol_Bypass (Failed)
	  5 - test_Unit_Aerosol_Bypass_TL (Failed)

With commit 5da55d3, (and after)

100% tests passed, 0 tests failed out of 162

Label Time Summary:
CRTM_Tests    = 186.58 sec*proc (162 tests)

Total Test time (real) = 189.63 sec

In 5da55d3, we made some changes to the selection of number of streams, and I modified the aerosol bypass ctests. The bugs and debug approaches were documented in PR #122.
I suspect what has been discussed in comment #122 (comment) are what causing the failed tests?