Question

Example

Given a number of different task types that need to be completed over a period of one month:

Type  |  Number per month
--------------------------
A     |  1
B     |  5
C     |  30
D     |  15
E     |  20
--------------------------
         71

Question:

How do I generate a flat order (1-dimensional array) of types that ensures that

• 71 jobs are completed within one month
• the types are spreaded as equally as possible (ABAB instead of AABB)

Edit:

One Idea that came up in the comments is:

Maybe reducing each stack of types by an amount relative to its size
is better than to use the absolute amount of 1. eg. A:2, B:10 results
in “for each A take 5 B” because B is five times larger than A.

Answer 1

Assume you have an array of task types such as:

types = [[:A,1],[:B,5],[:C,6],[:D,2],[:E,3]]

You can first sort, and then use the greater number as reference point to build the array of tasks.

Example:

types = [[:A,1],[:B,5],[:C,6],[:D,2],[:E,3]]
tasks=[]; i=0

asc = types.sort {|a,b| a[1] <=> b[1] }
max = asc[types.length-1][1]

(1..max).to_a.each do |i|
    types.each {|t| tasks << t[0] if t[1] >= i}
end

puts tasks.inspect

produces:

[:A, :B, :C, :D, :E, :B, :C, :D, :E, :B, :C, :E, :B, :C, :B, :C, :C]