How do you save benchmark results? about benchmarktools.jl HOT 2 CLOSED

results = run(suite, verbose = true)

BenchmarkTools.save("output.json", median(results))

cf. https://github.com/benchmark-action/github-action-benchmark/blob/master/examples/julia/fib.jl

seems to work and

BenchmarkTools.load("output.json")

also seems to work. So I guess these are just undocumented

from benchmarktools.jl.

I found the BenchmarkTools.save method too verbose for my liking. You can also save the benchmark results as follows:

using BenchmarkTools
using JSON

t = @benchmark fib(n) samples=10
data = Dict("mean"=> mean(t).time, "min"=> minimum(t).time, 
"max"=> maximum(t).time, "med"=>median(t).time)
open("output.json", "w") do f  JSON.print(f, data) end

Note: the time units are stored in nanoseconds. See: #307

from benchmarktools.jl.