Just a few benchmarks to compare concatenation (<>), interpolation, joining, and IO lists performances in Elixir

elixir_string_benchmarks.exs

Mix.install([:benchee])

defmodule StringPerf do
  def concat([]), do: ""
  def concat([head | tail]), do: head <> concat(tail)

  def interpol([]), do: ""
  def interpol([head | tail]), do: "#{head}#{interpol(tail)}"

  def joiner(input), do: Enum.join(input, "")

  def io(input), do: IO.iodata_to_binary(input)
end

# build just a list of stringified numbers
input = Enum.map(1..10_000, &Integer.to_string/1)

# check that all the functions from above return the same result
IO.inspect([
  StringPerf.concat(input) == StringPerf.interpol(input),
  StringPerf.interpol(input) == StringPerf.joiner(input),
  StringPerf.joiner(input) == StringPerf.io(input)
], label: "===> All functions return the same result")

# benchmark the performance!
Benchee.run(
  %{
    "concatenation" => fn -> StringPerf.concat(input) end,
    "interpol" => fn -> StringPerf.interpol(input) end,
    "join" => fn -> StringPerf.joiner(input) end,
    "io" => fn -> StringPerf.io(input) end
  },
  time: 5,
  memory_time: 5
)

results.txt

$ elixir string_bench.exs
===> All functions return the same result: [true, true, true]
Operating System: macOS
CPU Information: Apple M3 Pro
Number of Available Cores: 11
Available memory: 18 GB
Elixir 1.16.1
Erlang 26.2.1
JIT enabled: true

Benchmark suite executing with the following configuration:
warmup: 2 s
time: 5 s
memory time: 5 s
reduction time: 0 ns
parallel: 1
inputs: none specified
Estimated total run time: 48 s

Benchmarking concatenation ...
Benchmarking interpol ...
Benchmarking io ...
Benchmarking join ...
Calculating statistics...
Formatting results...

Name                    ips        average  deviation         median         99th %
io                  16.60 K      0.0603 ms     ±7.86%      0.0597 ms      0.0688 ms
join                 6.33 K       0.158 ms    ±14.11%       0.150 ms        0.22 ms
interpol             0.23 K        4.30 ms     ±7.55%        4.27 ms        5.41 ms
concatenation        0.23 K        4.37 ms     ±8.35%        4.29 ms        5.58 ms

Comparison:
io                  16.60 K
join                 6.33 K - 2.62x slower +0.0977 ms
interpol             0.23 K - 71.42x slower +4.24 ms
concatenation        0.23 K - 72.53x slower +4.31 ms

Memory usage statistics:

Name             Memory usage
io                  0.0469 KB
join                156.30 KB - 3334.33x memory usage +156.25 KB
interpol            859.29 KB - 18331.50x memory usage +859.24 KB
concatenation       859.29 KB - 18331.50x memory usage +859.24 KB

**All measurements for memory usage were the same**