Heiko Strathmann

Awesome initiative! I will check this out a bit more soon, first need to read up on Concepts I guess ;)

Reported as spam

Thanks for reporting! I think this should raise an assertion error, @lambday ? @ambodi it won't return anything sensible. The test statistic might be OK (depending on whether or not...

> @karlnapf it does not raise an error but returns nans. > > Great question, I have this case: I have a function that outputs one vector and another function...

Not sure this is still requires. Task is now to CHECK :)

currently, this is only supported without function arguments ``` watch_method("w_for_var", &LeastAngleRegression::get_w_for_var); ``` will not work out of the box. We could add this for functions with an index argument in...

ah but the dimensionality will be different or? If not, then the matrix should be fine. If so .... then we could add watch_method for methods with a single index...

I agree it is one way to solve the problem without having to twist the put/get/run api too much. For me this would be a temporary solution, where the proper...

We want something like this to work. ``` k = kernel("GaussianKernel", log_width=2) k.get("log_width") # returns 2 k.random_content() # randomizes all parameters k.get("log_width") # returns a the randomizes log width k.random_content("log_width")...

Shouhd be fixed in #3461 and the subsequent https://github.com/shogun-toolbox/shogun/commit/a39e8e4da4f3a71c1360b76c47e194548851326e