New python test framework is required.

raising as an issue.

Python needs random tests. See this

Please use new python test framework.

Let's close it for now. If I give it a try, one day, I'll reopen something if it still seems appropriate.

Ok, thank you.
Should I consider the problem as resolved, or come here in half a year and try to change it once again?

Hi

Here, you somehow enter the weird world of personal tastes. There is no real rules. Moreover, the feedback from the very same user can vary from one day to another...

Nothing more useful to say, sorry... :/

One more question... It's my first kata, I would like to know how to make it interesting.
I noticed that even though the problem is interesting, it's hard to present it in interesting way. Plenty of text, many parameters are the biggest issue here - at least I think so. I though that for people who like algorhitms it's a nice task, but it seems I was wrong.

Could you give me a tip, the idea, how would you create a task about epidemic and exponential growth?

As there is no reply and I think it's clear now what to do if multiplier is not integer, I'm marking issue as resolved.

I added a performance check. Let me know if it's a good idea, as numbers there are not realistic. On the other hand we may want to find parameters for the disaease and to do this we'll be running the model thousands of times. Maybe it's better to use possible values and just loop it?

Changed to: "In case multiplier is a fraction, round the result numbers only." Is it acceptable?

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.

You definitely shoudn't round each day, although I can understand why you consider it.
Generaly speaking for every single value here we take a mean for thousands of people. Doing it by looking at individuals we should look like this:

• Incubation time: 2-14 days
• Infecting time: 1-20 days
  Ofc we need to take these numbers randomly, set up weight. That's why it's better to simplify.

Anyway, there is an information that results must be rounded. Should I add any additional information?

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.

Thank you for pointing it out. I changed the definition of multiplier and missed it in that comment. Fixed.