Sum Optimized

Mix.install([
  {:benchee, "~> 1.1"}
])

Section

defmodule ListSum do
  def sum([]), do: 0
  def sum([head | tail]), do: head + sum(tail)

  def sum_optimized(list, accumulator \\ 0)
  def sum_optimized([], accumulator), do: accumulator
  def sum_optimized([head | tail], accumulator), do: sum_optimized(tail, head + accumulator)
end

Benchee.run(
  %{
    "sum" => fn input -> ListSum.sum(input) end,
    "sum_optimized" => fn input -> ListSum.sum_optimized(input) end,
    "control" => fn input -> Enum.sum(input) end
  },
  inputs: %{
    "large" => Enum.to_list(1..1_000_000),
    "medium" => Enum.to_list(1..10000),
    "small" => Enum.to_list(1..100)
  },
  time: 10,
  memory_time: 2
)

Operating System: Linux CPU Information: Intel(R) Core(TM) i5-10300H CPU @ 2.50GHz Number of Available Cores: 8 Available memory: 11.61 GB Elixir 1.13.3 Erlang 24.3

Benchmark suite executing with the following configuration: warmup: 2 s time: 10 s memory time: 2 s reduction time: 0 ns parallel: 8 inputs: list Estimated total run time: 28 s

Benchmarking sum with input list … Benchmarking sum_optimized with input list …

With input list

Name ips average deviation median 99th % sum_optimized 984.70 1.02 ms ±41.07% 0.94 ms 2.80 ms sum 299.00 3.34 ms ±29.06% 3.08 ms 7.56 ms

Comparison: sum_optimized 984.70 sum 299.00 - 3.29x slower +2.33 ms

Memory usage statistics:

Name Memory usage sum_optimized 600 B sum 600 B - 1.00x memory usage +0 B

All measurements for memory usage were the same