While looking into this I noticed another small issue with the calculations.

For the calculation of total_children_served_with_loose_disposables, we are doing division between integers so we are effectively rounding down for each item. So we could have 100 line items each with a quantity of 49, but if the distribution quantity of those items is 50, then we will get a total_children_served_with_loose_disposables value of 0.

We then use this sum to calculate the average monthly children served (converted to a float), so the average_children_monthly would be 0.

Is this expected behavior?

We probably should do a float division.

@cielf I will make another PR for that. Double checking these calculations in this PR is already complicated enough imo

@coalest Hrm. Yes. I don't care for the numbers that I'm currently getting for non-kitted diapers. All the "quantity per individual" is showing as blank for all, so with 12001 diapers, I would think I would get 240 total children served. I'm getting 235. So the float division is pretty important, actually -- Given the normal use, it's actually higher priority than getting the kits. right. But on with the kits...

I can throw the floating point change into this PR if you prefer.

I think what we want is to do the division as floating point, and then do a ceiling on that value -- at least for the total children served. A fractional child is a child for these purposes. Display total children helped as an integer.

The average children helped -- displaying that as a float (maybe 2 decimals) would make sense.

And yeah -- it will actually be easier, I think, to reason about this if we address the error in the 'loose' numbers.