What do you suggest?
To add a warning at the beginning of the documentation like xarray does (see http://xarray.pydata.org/en/stable/generated/xarray.ufuncs.angle.html#xarray.ufuncs.angle for examples) or to write our own documentation and examples for each ufunc?

Writing our own doc for all ufuncs would take too much time at this point. We could have an hybrid approach: write our own doc for the few ufuncs which are very important to our users and add a warning like xarray for the others. I consider the following ufuncs very important:

• where, minimum, maximum -- those needs to be both documented and have specific examples because broadcasting by axes names makes them more user-friendly than their numpy version.
• round, floor, ceil, trunc, isnan, isinf, sqrt, log, exp, inverse -- at least those need to appear in the API doc and ideally have our own examples.
• interp -- should appear in the API doc if it actually works with LArrays, which I have never tested (and I doubt is the case). Might be worth having specific documentation if it works well. My guess is that we should rather make a new LArray method for that one.
• all other ufuncs should have the numpy doc with a warning.

* round, floor, ceil, trunc, isnan, isinf, sqrt, log, exp, inverse -- at least those need to appear in the API doc and ideally have our own examples.

created new issue #715

* interp -- should appear in the API doc **if** it actually works with LArrays, which I have never tested (and I doubt is the case). Might be worth having specific documentation if it works well. My guess is that we should rather make a new LArray method for that one.

created new issue #716