As part of some Ruby-Processing work I’m doing, I needed a random number within a specified range. Processing has its own random function, but I wanted to use Ruby.

I needed a random number in the interval (-range, range). I immediately thought, “specify the max value range*2 as a param to rand, and then subtract range to get it into the interval -range to range.” My autopilot first attempt yielded:

    num = rand(range*2.0) - range

However, I didn’t read the documentation for Kernel#rand closely. According to the docs, the max parameter is converted to an integer:

    max1 = max.to_i.abs

This means that when I use a float less than zero as the range, say 0.25, it’s converted to 0. When zero is passed in, the range for rand is from 0 to 1.

All this makes sense, but I just didn’t think about it at first. Instead of passing in a max parameter, I needed to manipulate rand’s output to get what I wanted:

    num = rand*range*2.0 - range

That’s better.

In general, for a random number in the interval (min, max), you would use:

    num = rand*(max - min) - min

This post is tagged