Kevin Anderson

> What happens is that it doesn't let me import code from other files. Maybe I should change IDE to program in python. I am using Visual Studio Code. Do...

> Then, I see that some tests are not executed. What can we do in order to solve it? This was caused by an unrelated problem and is solved in...

> I think the reason is that it never occurs the case that "module type" and "Coefficients" are different from "None" in the test file. For this you can define...

> What about a third option - complement the existing iam.marion_diffuse with a pre-calculated lookup table and interpolation? Not wholly opposed, but the cool part of `marion_diffuse` is that it...

> Does anyone know why it says "TMYData" 70 times in the documentation? A long time ago in a commit far, far away, that's what the function's return value was...

> For now, I have just removed it from the file. I think technically that would be in violation of the CPR license file included in the zip archive posted...

> Minimally, static duck typing will provide a much clearer documentation of the expected type via a type annotation vs. a documentation comment which is subject to typos. > But...

> if we're going to go to all the effort of documenting it, why not just type it? My understanding is that neither PEP544 (Protocols) nor PEP484 (Type Hints) actually...

I think I'm against it? If we (1) don't want to be using `array_like` differently from numpy, AND (2) want output type to match input type, then I think any...

@adriesse I've been led to understand that this feature and #1523 are the first of several planned additions to pvlib. What do you think about opening (preliminary) issues for those...