Qué es Recursión

Clase 14 de 24 • Curso de Introducción a Elixir

Usuario anónimo

user•

Para hacer más corta las funciones anónimas se puede hacer así

Enum.map([1, 2, 3, 4, 5], &(&1 * 2))
# [2, 4, 6, 8, 10]

Donde &(...) es la función y &1 es el primer parámetro de la función

Juan David Gaviria Agudelo

student•

La función de parada en una sola linea:

def print_multiple_times(_msg, 0), do: :ok

Luis Enrique Hernandez Navarro

student•

en tiempo de ejecución, cuál seria más rápida?

print_multiple_time(_msg, n) when n == 0, do: :ok

print_multiple_time(_msg, 0),  do: :ok

Milton Mazzarri

teacher•

Ambos modos son bastantes rápidos (en el orden de los nano segundos en una prueba rápida que hice). Si bien te diría que es muy temprano para preocuparte por el performance, te explico como puedes obtener algunos datos que podrían ayudarte a tomar una decisión en caso de ser necesario en el futuro.

Usando el paquete benchee y teniendo lo siguiente:

# lib/demo.ex
defmodule Demo do
  def first(_msg, n) when n == 0, do: :ok

  def second(_msg, 0), do: :ok

  def third(_msg, n) when n === 0, do: :ok
end

Nota que introduje un tercera función llamada third, en este caso en la guarda uso el operador de comparación estricto de Elixir, lo cual lo haría equivalente a la función second.

Con este pequeño script:

# bench.exs
Benchee.run(
  %{
    "first" => fn -> Demo.first(:ok, 0) end,
    "second" => fn -> Demo.second(:ok, 0) end,
    "third" => fn -> Demo.third(:ok, 0) end
  },
  time: 10,
  memory_time: 2
)

Puedes ver que tanto second y third son equivalentes en el número de millones de operaciones que procesan.

mix run bench.exs
Operating System: macOS
CPU Information: Apple M1 Max
Number of Available Cores: 10
Available memory: 64 GB
Elixir 1.13.4
Erlang 25.0.2

Benchmark suite executing with the following configuration:
warmup: 2 s
time: 10 s
memory time: 2 s
reduction time: 0 ns
parallel: 1
inputs: none specified
Estimated total run time: 42 s

Benchmarking first ...
Warning: The function you are trying to benchmark is super fast, making measurements more unreliable!
This holds especially true for memory measurements or when running with hooks.

See: https://github.com/bencheeorg/benchee/wiki/Benchee-Warnings#fast-execution-warning

You may disable this warning by passing print: [fast_warning: false] as configuration options.

Benchmarking second ...
Warning: The function you are trying to benchmark is super fast, making measurements more unreliable!
This holds especially true for memory measurements or when running with hooks.

See: https://github.com/bencheeorg/benchee/wiki/Benchee-Warnings#fast-execution-warning

You may disable this warning by passing print: [fast_warning: false] as configuration options.

Benchmarking third ...
Warning: The function you are trying to benchmark is super fast, making measurements more unreliable!
This holds especially true for memory measurements or when running with hooks.

See: https://github.com/bencheeorg/benchee/wiki/Benchee-Warnings#fast-execution-warning

You may disable this warning by passing print: [fast_warning: false] as configuration options.


Name             ips        average  deviation         median         99th %
second      278.89 M        3.59 ns   ±237.57%        3.58 ns        3.71 ns
third       278.84 M        3.59 ns   ±206.41%        3.58 ns        3.71 ns
first       159.99 M        6.25 ns ±20028.18%        4.20 ns        8.40 ns

Comparison:
second      278.89 M
third       278.84 M - 1.00x slower +0.00066 ns
first       159.99 M - 1.74x slower +2.66 ns

Memory usage statistics:

Name      Memory usage
second             0 B
third              0 B - 1.00x memory usage +0 B
first              0 B - 1.00x memory usage +0 B

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

Espero esto responda tu inquietud.

# bench.exs
Benchee.run(
  %{
    "first" => fn -> Demo.first(:ok, 0) end,
    "second" => fn -> Demo.second(:ok, 0) end,
    "third" => fn -> Demo.third(:ok, 0) end
  },
  time: 10,
  memory_time: 2
)

mix run bench.exs
Operating System: macOS
CPU Information: Apple M1 Max
Number of Available Cores: 10
Available memory: 64 GB
Elixir 1.13.4
Erlang 25.0.2

Benchmark suite executing with the following configuration:
warmup: 2 s
time: 10 s
memory time: 2 s
reduction time: 0 ns
parallel: 1
inputs: none specified
Estimated total run time: 42 s

Benchmarking first ...
Warning: The function you are trying to benchmark is super fast, making measurements more unreliable!
This holds especially true for memory measurements or when running with hooks.

See: https://github.com/bencheeorg/benchee/wiki/Benchee-Warnings#fast-execution-warning

You may disable this warning by passing print: [fast_warning: false] as configuration options.

Benchmarking second ...
Warning: The function you are trying to benchmark is super fast, making measurements more unreliable!
This holds especially true for memory measurements or when running with hooks.

See: https://github.com/bencheeorg/benchee/wiki/Benchee-Warnings#fast-execution-warning

You may disable this warning by passing print: [fast_warning: false] as configuration options.

Benchmarking third ...
Warning: The function you are trying to benchmark is super fast, making measurements more unreliable!
This holds especially true for memory measurements or when running with hooks.

See: https://github.com/bencheeorg/benchee/wiki/Benchee-Warnings#fast-execution-warning

You may disable this warning by passing print: [fast_warning: false] as configuration options.


Name             ips        average  deviation         median         99th %
second      278.89 M        3.59 ns   ±237.57%        3.58 ns        3.71 ns
third       278.84 M        3.59 ns   ±206.41%        3.58 ns        3.71 ns
first       159.99 M        6.25 ns ±20028.18%        4.20 ns        8.40 ns

Comparison:
second      278.89 M
third       278.84 M - 1.00x slower +0.00066 ns
first       159.99 M - 1.74x slower +2.66 ns

Memory usage statistics:

Name      Memory usage
second             0 B
third              0 B - 1.00x memory usage +0 B
first              0 B - 1.00x memory usage +0 B

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

Qué es Recursión

Conceptos básicos

Programación Funcional en Elixir para Backend

¿Por qué Elixir?

Primeros pasos con Elixir

Introducción a Livebook

Tipos de datos básicos: Operaciones aritméticas, booleanos, átomos, y cadenas

Tipos de datos básicos: funciones anónimas, listas, y tuplas

Introducción al lenguaje Elixir

Operadores básicos

Descomponer Estructuras de datos via Pattern Matching

Explorando estructuras de control: cond, else, with

Explorando estructuras de control: If y unless

Keyword lists, maps

Conociendo las funcionalidades de Elixir

Módulos en Elixir

Funciones en Elixir

Qué es Recursión

Enumerables

Flujos o Streams

Qué son los Procesos

Módulos: IO, File, Path

Alias, require, import

Anidamiento de módulos y multi alias/require/import

Atributos de un módulo

Proyecto

Entendiendo las reglas del juego, primeros pasos

Bonus

Uso de Mix

Cierre

¿Qué sigue?