Erik Tollerud

Document and test "unpacking" of compound models in fitters

1

See https://github.com/astropy/specutils/pull/210#discussion_r215767298 . I'm not sure how best to document this - perhaps as a worked example in the just-merged #210 fitting narrative docs?

From #210 - see https://github.com/astropy/specutils/pull/210#discussion_r214480177 for more detail. The basic gist: models like polynomials can't be "copied" trivially because you need a way to know how many parameters they have....

Prompted by https://github.com/astropy/specutils/pull/210#discussion_r214478583 : right now the "window" functionality of the fitting machinery assumes the model has a `center`. Many models don't have such a parameter - it goes by...

Make some "science use case" tests for the analysis tools

1

Right now some of the unit tests do things that aren't always that meaningful scientifically, like measuring the centroid of multiple lines or the like. That is often fine/good for...

Option to state which SHA to compare to

1

Right now it appears (at least with Travis) that coveralls always does comparisons against master when deciding how much coverage increases/decreases. This is problematic on an active project where the...