The first step to understanding anything is to measure it. The same goes for writing high-performance Julia code; we need to measure the performance of the code as the first step to achieving this. Fortunately Julia makes this extremely easy for us. There are simple ways to measure the time taken by any Julia code built into the Julia runtime. Moreover, if you want to perform statistically accurate benchmarking, there are high-quality packages available.
The simplest way to measure time in Julia is using the tic()
and toc()
functions. Place these functions respectively before and after any piece of Julia code, and we will note the time taken by this code on the console. Run the following code:
julia> tic(); sqrt(rand(1000)); toc(); elapsed time: 0.000137693 seconds
In the preceding code, we measured the time taken to generate 1,000 random numbers, and to compute its square root. Technically, all the toc()
function does is print the elapsed time
value since the...