How to optimize the program with two for loops - ruby-on-rails

Question

I have a following programm
def calc_res(a)
n = a.length
result = 0
for i in 0 .. (n - 1)
for j in i .. (n - 1)
if (a[i] != a[j] && j - i > result) then
result = j - i
end
end
end
return result
end
which return following output
irb(main):013:0> calc_res([4, 6, 2, 2, 6, 6, 4])
=> 5
but it is taking time if array size is too large e.g. [0,1,2,3,.....70000]
can any one suggest me how can I optimize it.
Thanks

Answer 1

If I have understood the problem you are trying to solve (from code)
def calc_res(a)
last_index = a.length - 1
index = 0
while a[index] == a.last do
index = index + 1
break if index == last_index
end
last_index - index
end
It checks items from start if they are equal to items from end, end it moves the index toward the last element. As I understood you search for max length between different elements.
For you problem with [4, 6, 2, 2, 6, 6, 4] it will have one iteration and return 5, for the problem with [1...70000] it will have zero iterations and will return the difference in positions for those two (size of the array - 1)

Answer 2

My understanding is that the problem is to find two unique elements in the array whose distance apart (difference in indices) is maximum, and to return the distance they are apart. I return nil if all elements are the same.
My solution attempts to minimize the numbers of pairs of elements that must be examined before an optimal solution is identified. For the example given in the question only two pairs of elements need be considered.
def calc_res(a)
sz = a.size-1
sz.downto(2).find { |n| (0..sz-n).any? { |i| a[i] != a[i+n] } }
end
a = [4,6,2,2,6,6,4]
calc_res a
#=> 5
If sz = a.size-1, sz is the greatest possible distance two elements can be apart. If, for example, a = [1,2,3,4], sz = 3, which is the number of positions 1 and 4 are apart.
For a, sz = a.size-1 #=> 6. I first determine if any pair of elements that are n = sz positions apart are unique. [a[0], a[6]] #=> [4,4] is the only pair of elements 6 positions apart. Since they are not unique I reduce n by one (to 5) and examine all pairs of elements n positions apart, looking for one whose elements are unique. There are two pairs 5 positions apart: [a[0], a[5]] #=> [4,6] and [a[1], a[6]] #=> [6,4]. Both of these meet the test, so we are finished, and return n #=> 5. In fact we are finished after testing the first of these two pairs. Had neither these pairs contained unique values n would have been reduced by 1 to 4 and the three pairs [a[0], a[4]] #=> [4,6], [a[1], a[5]] #=> [6,6] and [a[2], a[6]] #=> [2,6] would have been searched for one with unique values, and so on.
See Integer#downto, Enumerable#find and Enumerable#any?.

Answer 3

A more rubyesque versions include:
def calc_res(a)
last = a.last
idx = a.find_index {|e| e != last }&.+(1) || a.size
a.size - idx
end
def calc_res(a)
last = a.last
a.size - a.each.with_index(1).detect(->{[a.size]}) {|e,_| e != last }.last
end
def calc_res(a)
last = a.last
a.reduce(a.size) do |memo, e|
return memo unless e == last
memo -= 1
end
end
def calc_res(a)
return 0 if b = a.uniq and b.size == 1
a.size - a.index(b[-1]).+(1)
end