Christopher Dupuis

@jhamman Ah, yeah that is a little unclear. `sc` is just a struct with lists of variable names. Maybe instead, you can think of it like this: model.fit([batch[x] for x...

No luck so far.

This style of warning suppression seems promising, but it's not working: import warnings def fxn(): warnings.warn(Warning) with warnings.catch_warnings(): warnings.simplefilter("ignore") fxn()

Adding `clear_output(wait=True)` in the `for batch in bgen` loop helps. It would make sense to add this to the documentation, I expect this will be a common problem. I'll see...

@jhamman I don't have a simple example of this handy, can I send you a gist of what I have?

I think I have something workable... Using my update to xbatcher [here](https://github.com/cmdupuis3/xbatcher/tree/batch_dim), you can run this: ```python import numpy as np import xarray as xr from IPython.display import clear_output import...

Taking...

I guess so? Someone told me to try working on this at the SciPy hackathon as a good next-step issue. Looks like the links here are out of date. Do...

I think that's similar to what I was saying, along the lines of separating your graph operations from your data operations. If we're just looking at the primitives (rather than...

I'm looking at this again, and I think the way to understand my PoV is that I want the least friction between the `for batch in bgen` line and where...