binary search. `elixir search.exs` to run this. assuming you have elixir (`brew install elixir`)

search.exs

defmodule Search do
  def chop(_val, []), do: -1

  def chop(val, [head|tail]) when tail == [] do
    if val == head, do: 0, else: -1
  end

  def chop(val, list) do
    chop(val, 0, length(list) - 1, list)
  end

  def chop(val, left, right, list) when right - left == 1 do
    cond do
      Enum.at(list, left) == val -> left
      Enum.at(list, right) == val -> right
      true -> -1
    end
  end

  def chop(val, left, right, list) do
    middle = round((right - left)/2)
    mid_val = Enum.at(list, middle)

    if (mid_val > val),
      do: chop(val, left, middle, list),
    else: chop(val, middle, right, list)
  end
end

test_search.exs

Code.require_file "../search.exs", __FILE__

ExUnit.start

defmodule TestChop do
  use ExUnit.Case

  test "chop on small or empty lists" do
    assert Search.chop(1, []) == -1
    assert Search.chop(1, [3]) == -1
    assert Search.chop(1, [1]) == 0
  end

  test "chop on normal lists" do
    assert Search.chop(1, [1, 3]) == 0

    assert Search.chop(1, [1, 3, 5]) == 0
    assert Search.chop(3, [1, 3, 5]) == 1
    assert Search.chop(5, [1, 3, 5]) == 2
    assert Search.chop(2, [1, 3, 5]) == -1
    assert Search.chop(6, [1, 3, 5]) == -1
  end

  test "more cases" do
    assert Search.chop(1, [1, 3, 5, 7]) == 0
    assert Search.chop(5, [1, 3, 5, 7]) == 2
    assert Search.chop(3, [1, 3, 5, 7]) == 1
    assert Search.chop(7, [1, 3, 5, 7]) == 3
    assert Search.chop(0, [1, 3, 5, 7]) == -1
    assert Search.chop(2, [1, 3, 5, 7]) == -1
    assert Search.chop(4, [1, 3, 5, 7]) == -1
    assert Search.chop(6, [1, 3, 5, 7]) == -1
    assert Search.chop(8, [1, 3, 5, 7]) == -1
  end
end

The catch here is that Enum.at works in linear time, so this has worse performance than just doing a linear search