Execution time and measurement

Intermediate Julia

Anthony Markham

Quantitative Developer

Initial function definition

While problems have multiple solutions, there is often an 'optimal' solution
Large memory usage and long execution times can frustrate both users and developers
Loops are a common cause of longer execution times and inefficiency

function my_function()
  x = Vector{Char}()
  for i in "Anthony"
    push!(x, i)
  end
  println(x)
end

['A', 'n', 't', 'h', 'o', 'n', 'y']

Advantages:

Disadvantages:

@time my_function()

@time my_function()

0.278266 seconds (110.66 k allocations: 6.343 MiB, 99.79% compilation time)

0.000493 seconds (100 allocations: 3.281 KiB)

Why use @benchmark over @time?

The most flexible option, lots of parameters to tweak, such as number of samples.
In-depth runtime statistics (range, median, mean).

@benchmark my_function

Intermediate Julia