Open itrharrison opened 3 years ago
Separate from data included in this repo, there is also test data for some likelihoods that gets pulled in via the InstallableLikelihood
functionality, e.g. for lensing here:
and for MFLike as well. So any pattern that tries to address simulated test data in a unified way needs to include those, too.
We also discussed another place which should switch to sacc. In the multi-Gaussian likelihood, which currently assumes Cls stored as .npy files:
At the moment there are a number of tests which are run using data files from different simulations. See the list below.
Eventually these simulations should be as coherent as possible.
For now I will check which of these data files are actually used and collect them together into a single
sacc
file which is used by the tests.The list of data files I can see are: