one on the expected answers with c has a 1 all the way in the 10thousandths place.

Enabled in this fork

Also description is inconsistent here:

These are for all languages except JavaScript

• Make sure to round off the result to two decimal points.

This is for JavaScript

• Your answer should be accurate within 0.01 kpl.

All-in-all, the rounding requirement should be removed for all languages (in-accordance with issue below) and replaced with the description in JavaScript

All languages should follow Java by using approximate equality comparison when writing custom tests

fork approved, issue fixed

That's because 2.35 is the conversion factor from mpg to kpl of a US gallon. The kata is specifically asking about imperial gallons, which are larger than US gallons. It's better to just use the numbers provided in the kata rather than search online for a conversion factor anyway, otherwise you are likely to run into rounding errors.

in python, this doesn't work...

return mpg / 2.35

10 - 4.25 (rounded) - expected 3.54

typed on calculator, got 4.25 (again rounded)

the problem comes from your code. you do not use the appropriate datatypes (note the return type)

problem with big numbers at random test on C.

Mathematics are correct, idk what wrong with them. Example:

Submitted: 7.636e+07

Expected: 7.64e+07

Задание чтобы вспоминить цикл for, создание листа и добавление в него

Print the input to help you debug your code, but your problem is about how you set the initial value of small2. Read this, please: https://docs.codewars.com/training/troubleshooting