Adding testing and validation

[arthir]

Various changes:

• Moving experiment config yamls into their own folder
• Added unit tests for function which processes the parameter function description
• Modified the date->timestep processing to work off an existing column

[arthir]

@stephen-hoover I added some basic tests for the parsing (and this also sets up a basic testing framework).

[stephen-hoover]

This all LGTM!

Some comments on things related to this PR that I've found useful:

• There's a pd.testing.assert_frame_equal function which is great for testing changes to DataFrames. Similarly, pd.testing.assert_series_equal.
• I find set literals easier to read than function calls, set([10]) == {10}.
• For testing with random numbers, I'd suggest creating a np.random.Generator object with controlled seed in the test function. That would require modifying add_config_parameter_column to accept an optional Generator parameter, which seems out of scope for this PR.

[arthir]

• There's a pd.testing.assert_frame_equal function which is great for testing changes to DataFrames. Similarly, pd.testing.assert_series_equal.
  Of course. I should have thought of this instead of the long winded way I was testing it.

• I find set literals easier to read than function calls, set([10]) == {10}.
  Makes sense - I think I removed all of these anyways.

• For testing with random numbers, I'd suggest creating a np.random.Generator object with controlled seed in the test function. That would require modifying add_config_parameter_column to accept an optional Generator parameter, which seems out of scope for this PR.
  Ah, that's a good point but yes, I think out of scope for now.

[stephen-hoover]

LGTM!

[stephen-hoover]

Neat! I didn't know about the match kwarg. That's useful.

[arthir]

Josh told me about this one today -- it's so nice!