1D responces using PSF model

[nocturnalastro]

If a rsp is supplied then the model is wrapped in a sherpa PSF model. This is only done to the sherpa fit model and not the input/returned astropy model.

1D and 2D rsp's are the same but this branch has an statement which makes the rsp=None if the dataset is 2D and the 2D

[hamogu]

There is clearly a lot going on in this PR. Is it intended that all the generated images are included in the git repro? Won't they be generated by rtd when the docs are build?

[nocturnalastro]

These are the all the commits I'll squash them today and no they shouldn't be there i'll remove them

[nocturnalastro]

I think it would be easier to pull master and rewrite the code than trying to fix this with a rebase

[olaurino]

I think it would be easier to pull master and rewrite the code than trying to fix this with a rebase

I am not sure I follow. Could you be more specific?

[nocturnalastro]

@olaurino fixing the commits will probably take longer than just writting the code from what exists in master and remaking the branch

[olaurino]

You can easily squash all commits, then you can fix any conflicts and then, if you want, you can split all the code in cleaner commits or in different branches if that's what you want. I am not sure what you mean by "fixing" though, so maybe I am missing something. Please feel free to skype me if you want to talk about it.

[nocturnalastro]

There's a lot more commits here than there should be, not sure how they ended up here.
I'm just thinking it will take me less time to rewrite the code than solving all the conflicts, each time rebase finds one. Most of the conflicts we're pretty easy to resolve. I don't think my git-fu is quite up to snuff yet.

[olaurino]

Which is why this might be a good time to improve your git-fu ;)

Rather than rebasing, simply squash all commits first in the current branch into a single commit. Then do a diff, and see whether what you see is more or less what you would like to get. Then start a new branch off master and cherry-pick the squashed commit (with --no-commit or something similar). Now it's as if you were editing new code. You can either commit everything or use git add -p to select which chunks to include in individual commits.

[nocturnalastro]

Cheers @olaurino It looks much nicer now :)
I missed the keyword arg ... so close

[coveralls]

Coverage decreased (-1.9%) to 24.638% when pulling 00fc661 on 1Dpsf into 7a75779 on master.

[olaurino]

Regarding coverage, have you tried adding --cov-report=xml to the arguments of the test command?