javascript translation posted.

Just like the JS and Haskell versions. I'll see what I can come up with though.
EDIT: I've added random tests. If acceptable I can add a similar set to the JS version, but somebody else will have to handle Haskell.

Python translation submitted

Oof, that's embarassing! Thank you for marking as resolved!

Please explain the meaning of cells labeled 'C' or remove these cells from the tests. This value is not mentioned in the instructions, and I can not intuit its behavior from the test cases.

You should add precisely what you told me - that fractional people are to be carried forward until the very end and only the returned results are to be rounded. That will inform solutions from the start instead of forcing a complete solution redesign.

I don't know. Should I be maintaining fractional people throughout the process? Or should I be rounding at each day? Because I feel like I should be rounding at each day since fractional people can't be infected. It's a binary state.

I'm confused by your rounding.

In the fifth example case you have a multiplier of 2.5. Assuming this means each infected person infects 2.5 new people per day should we be using math.floor (2.5 yields 2), math.ceil(2.5 yields 2), or round (2.5 yields 2) to determine how many new incubating cases there are when there is 1 infecting person on day 4?

I'm confused by your multipler.
In the instructions you say that a multiplier of 1 means that each infecting person infects one new person each day.
But in the test cases you say that a multiplier of 2 means that each infecting person infects one new person each day.
Which is it, and how is the multiplier intended to work?

Description should list expected errors to be thrown instead of making the developer fail hidden test cases over and over getting new errors each time.

Tests need to fail when output array contains undefined rather than any of the house names in the param array.