Remove duplicates from an array of array

As previously pointed out

product_areas = [["1", "2", "3"], ["3", "1", "2"]].flatten.uniq

-OR-

product_areas.flatten.uniq!

will both lead you to your desired answer.

Why?

When you were running "product_areas.uniq!" the process was comparing the two inner arrays against each other, other than the elements of each array. Because both ["1", "2", "3"] and ["3", "1", "2"] are unique in the array, neither will be removed. As an example say you had the following array

product_areas = [["1", "2", "3"], ["3", "1", "2"], ["1","2","3"]]

and you ran:

product_areas = product_areas.uniq

product_areas would then look like the following:

product_areas = [["1", "2", "3"], ["3", "1", "2"]]

What you need to be aware of when running any sort of enumerable method on arrays is it will only move down to each individual element. So if inside an array you have more arrays, any iterative method will look at the inner array as a whole. Some sample code to demonstrate this:

array_of_arrays = [[1,2,3], [4,5,6]]

array_of_arrays.each do |array|
  p array
end
#---OUPUT---
# [1, 2, 3]
# [4, 5, 6]

array_of_arrays.each do |array|
  array.each do |element|
    p element
  end
end
#---OUPUT---
# 1
# 2
# 3
# 4
# 5
# 6

I've used this little snippet of code over and over again in my career as a Ruby on Rails developer to solve this frequently encountered problem in a single piece of neat little code. The end result of this code is that you can just call something like

product_areas.squish

to do the following:

1. flatten 2-D arrays into 1-D arrays
2. remove nil elements
3. remove duplicate elements

I do this by adding an initializer config/initializer/core_ext.rb to my project which extends the core Ruby functionality as follows:

class Array
  def squish
      blank? ? [] : flatten.compact.uniq
  end
end

class NilClass
  def squish
      []
  end
end