Alexander Condello

I'll mark it as both! :smile:

One potential alternative to the attribute pattern would be something like ```python import abc import numpy as np class BQM_float64: pass class BQM_float32: pass class BQM(abc.ABC): def __new__(cls, *args, dtype=np.float64,...

Can you please post a [minimal reproducible example](https://stackoverflow.com/help/minimal-reproducible-example)? The following works fine for me ```python In [3]: import dimod In [4]: i = dimod.Integer('k_0', lower_bound=0, upper_bound=38) In [5]: i /...

I agree that in the context of symbolic manipulation, this might be a bit unexpected. The reason that we keep them is because there are several ways to conceptualize the...

> To clarify, I think the order should not change, just an intermediate label would be removed (for example) - no relabeling of the nodes should happen. Perhaps they would...

Another option would be some sort of global or class-level toggle for this behavior. That would allow us to gracefully deprecate the existing behavior, if we did decide to do...

> To clarify, I was thinking of removing edges, not variables. But it's a valid point - if you remove all the edges that contain a variable, shouldn't you remove...

As a downside, it does not seem like those type annotations are picked up by sphinx (see https://github.com/sphinx-doc/sphinx/issues/7630).

I am a tiny bit hesitant to implement some of these, pending a resolution to https://github.com/dwavesystems/dimod/issues/849. Some useful snippets in the mean time ```python num_variables = len(cqm.variables) num_binaries = sum(cqm.vartype(v)...

Hi @AkashNarayanan , I agree. I think there are a few things that should be done - [ ] Add a docstring - [ ] Add the method [to the...