Not sure, just trained this kata. I suggested to add a testcase for this.

It would be great to add a testcase for test_even(2.0)

This would cover the float case better. Many submitted solutions pass the existing tests where they should fail.

It shouldn't!

[1] pry(main)> def test_even(n)
[1] pry(main)*   n % 2 == 0
[1] pry(main)* end
=> :test_even
[2] pry(main)> test_even(2.0)
=> true
[3] pry(main)> 2.0 % 2
=> 0.0
[4] pry(main)> 0.0 == 0
=> true

Maybe you're thinking of .eql?:

[5] pry(main)> 0.0.eql? 0
=> false

Edit: just observed that floats should be considered uneven for this, so I eat my words! Has the kata changed since it was originally written?

This would fail test_even(2.0)

I have to say it is a ugly feature of ruby to use % for formatting. I would rather use sprintf or other helper methods to achieve this.

I'm curious as to how this works. Any links to what the percent operator does in this case? Whatever it is, it looks fancy and useful.
EDIT: Found it after some tough google work
http://www.ruby-doc.org/core-2.1.1/String.html#method-i-25

The input string needs to be defined more precisely. Input validation needed?