Fix calibration errors in multiband likelihood

This MR fixes a bug in multiplying calibration errors in multiband likelihood. This also adds unit tests of multiband likelihood for non-zero calibration errors, so that such mistakes will never occur in the future.

added Bug Likelihood labels

assigned to @soichiro.morisaki

unassigned @soichiro.morisaki

approved this merge request

Thanks @soichiro.morisaki, just for the record, I ran the new tests with the latest release and three of the new jobs failed, so the new code has definitely fixed the behaviour.

========================================================================= short test summary info ==========================================================================
FAILED ../../../test/gw/likelihood_test.py::TestMBLikelihood::test_homs_1 - AssertionError: 16.152660489949767 not less than 0.001
FAILED ../../../test/gw/likelihood_test.py::TestMBLikelihood::test_ifft_fft_1 - AssertionError: 16.147603120524764 not less than 0.006
FAILED ../../../test/gw/likelihood_test.py::TestMBLikelihood::test_matches_non_mb_1 - AssertionError: 16.142027530970836 not less than 0.015
========================================================= 3 failed, 4 passed, 183 deselected, 1 warning in 10.91s ==========================================================

approved this merge request

merged

mentioned in commit c629e770

changed milestone to %1.2.0