scott bridwell

I would say that stating specific column dependencies, while cumbersome, is still preferable to having to manually clear the cache within the steps. But maybe this could be automated in...

I wrote a simple context manager that manages the numpy and python random states. Might be applicable here: https://github.com/AZMAG/smartpy_core/blob/master/smartpy_core/wrangling.py#L12 https://github.com/AZMAG/smartpy_core/blob/master/smartpy_core/tests/test_wrangling.py#L163

I did some playing around w/ this, if we update the init in _ModelFuncWrapper from: ``` def __init__(self, model_name, func): self.name = model_name self._func = func self._arg_list = set(inspect.getargspec(func).args) ```...

I think the current behavior is what is expected/desired. For example, we re-add (essentially overwrite) the households table each time the household transition model is run. So we still need...

This might be a short term fix, but how's something like this? Basically define a step that will change the scope of the column at the start of each year....

Cool, I think we're in the same session in Denver. I'll be talking about some of my ideas for orca/caching issues as well. Should be an interesting discussion!

@fscottfoti, yes to all of the above. I haven't really worked with orca_test yet, but you're right, there does seem to be some overlap and perhaps that would be a...

The more I look at this, I feel like the whole expressions approach should be moved out of the function defintion and into the step/injectable/table/column definition. To me, it's kind...

I'm fine with changing the default. Or maybe, the default caching scope could be be a global option?

I like the suggestion of emulating the travel direction with another impedance column. The only issue might be for true one-way streets that can only be traveled in one direction....