Small changes to allow IC to use MixedChiSquare; also a quick hack to…

[stewartsiu]

… ChiSquared test to allow zero contingency entry, so that it can be used for testing.

[akelleh]

Hey @stewartsiu ! We shouldn't have print statements outside of exceptions!

[akelleh]

Also, line 32 needs a variable_types={} kwarg!

[genemiller]

@akelleh, Why line 32 (it is a class definition)? If you meant line 33, then wouldn't additional code be needed to handle variable_types?

[akelleh]

you're right! line 33. No additional code needs to be added to handle it. It's required since the ICs.search() method passes the CIT a variable_types arg by default now

[akelleh]

@stewartsiu (so the modification on line 33 should be all we need!)

[stewartsiu]

Hi @akelleh!

1. On the print statement, what is your preferred mechanism of displaying info about the independence test? I just added that line when I was debugging to help myself see what it's working on.
2. On line 33, I didn't change it because I wasn't sure if you intend ChiSquaredTest to be a standalone independence test (I seem to remember some issues with data format) or merely as a helper method for MixedChiSquaredTest.

[akelleh]

Hey @stewartsiu
(1) It's fine to print in general, but I want to avoid adding any more print statements to master, unless they're reporting errors. We want this to be something that doesn't dump a bunch of text to stdout, which is generally bad practice (unless you call a method that specifically asks for text).

(2) That's cool. It'll work as a CIT on discrete data, so we need to make sure it's still compatible with ICs, so we need to make sure it accepts the args that ICs passes to it. Let me know if you have any questions!