Should 3sat be in examples rather than experiments? about tensornetwork HOT 10 CLOSED

Then again, it doesn't currently have a script component that runs an instance. Could it acquire one? :)

from tensornetwork.

I guess the bigger question is what difference should exist between experiments/examples? Should there even be one?

from tensornetwork.

That is a bigger question. It seemed sensible to distinguish when @mganahl and I were discussing this with @sleichen, since there is a big difference in code complexity. However, I suppose we could just add READMEs for each one explaining what it is and indicating its complexity (or suitability as a pedagogical example for beginners).

from tensornetwork.

Hmm. How about we merge them into one folder call extensions since they basically all extend TensorNetwork functionality.

from tensornetwork.

I think examples should contain a collection of code highlighting basic use cases of the API, without too much additional stuff happening around it, so that users can get started quickly with the basic things. The experiments folder for now contains relatively heavy weight numerics, and I don't think most of the stuff there is of any use for someone who just wants to see the API in action. So as long as we don't have a more extensive documentation, I'd put stuff which uses the API directly, and isn't too complicated to understand, in the examples folder. I agree that we could merge the two if we addedd appropriate documentation and the like

from tensornetwork.

Yeah that seems reasonable to me.

Ok cool, examples will be for teaching and experiments will be for complicated stuff that we might make part of the library in the future. SGTM.

Now, should 3SAT be an example or an experiment?

from tensornetwork.

I'd put it in examples. It's not too complicated, and its a good example of how to use the API

from tensornetwork.

At the moment I agree. We can still decide to make it a part of the API later, if we want.

from tensornetwork.

Cool. Anyone wanna make a PR? 😄

from tensornetwork.

#49

from tensornetwork.