How does Ruby's map method work in this case? - ruby-on-rails

Question

I got the mistake when I want to add doubled values to an array:
arr = [1,2,3]
def my_mistake(arr)
result = Array.new
arr.map { |element| result << element * 2 }
end
#=> [[2, 4, 6], [2, 4, 6], [2, 4, 6]]
def solution(arr)
arr.map { |element| element * 2 }
end
#=> [2,4,6]
However, come back to my mistake and the definition of map method in Ruby.
Invokes the given block once for each element of self. Creates a new array containing the values returned by the block.
I think my_mistake method has to return [[2], [2, 4], [2, 4, 6]] but it doesn't.
Everyone can explain this case for me ?

Answer 1

The resulting array will contain three occurrences of the same reference to the same array, as result is the result of the expression result << element * 2. So the result of the map is (kind of) [result, result, result]. These all point to the same content, which is the content of result at the end of the process ([2, 4, 6]).
What you expected would be achieved if you clone the array at each point, so that every resulting element would point to a different array, and each addition would not affect the previously computed arrays:
arr.map { |element| (result << element * 2).clone }
=> [[2], [2, 4], [2, 4, 6]]

Answer 2

.map returns the last evaluated expression, so no need for the result << part there. Here's something that worked for me:
def my_mistake(arr)
result = [] # '= []' is same like '= Array.new', look-up "literal constructors in Ruby"
new_arr = [] # same like new_arr = Array.new
until arr.empty?
new_arr << arr.shift # we add each element of arr, one by one, starting from the beginning
output << new_arr.map { |e| e * 2} # we calculate *2 for each element
end
return result
end
p my_mistake(arr) #=> [[2], [2, 4], [2, 4, 6]]
If you're nto sure how this works, try putting "p output" after the 6th line:
def my_mistake(arr)
output = []
new_arr = []
until arr.empty?
new_arr << arr.shift
output << new_arr.map { |e| e * 2}
p output
end
return output
end
my_mistake(arr)
The program will print:
[[2]]
[[2], [2, 4]]
[[2], [2, 4], [2, 4, 6]]