re: https://twitter.com/SimonDeDeo/status/1344031423704006656
You are not timing the same thing in each language.
See below for what you were timing in Julia.
And, as has been mentioned, use @time or @btime from BenchmarkTools.jl
(time does not do what you expect it to do).
When using @time, run the timing once to precompile the test
before considering the results valid. The first run may take
considerably longer, and that extra time is not computing time.
I have omitted those lines below.
# this gets transformed into an array of Float64s
julia> bb = [1, 2.0]
2-element Array{Float64,1}:
1.0
2.0
# so, here we are timing promotions, allocations, and assignment
julia> @time for i in 1:100000; b=[1,2.0]; end
0.150965 seconds (400.00 k allocations: 16.785 MiB)
# here we are timing allocations, and assignment
julia> @time for i in 1:100000; b=[1.0,2.0]; end
0.002405 seconds (100.00 k allocations: 9.155 MiB)
# here we are timing the assignment only
julia> bb = [1, 2.0];
julia> @time for i in 1:100000; b=bb; end
0.000239 seconds # no allocations
# You can work with mixed type arrays,
# to do that efficiently requires explict typing
julia> bb = Union{Int,Float64}[1, 2.0]
2-element Array{Union{Float64, Int64},1}:
1
2.0
# here we are timing the assignment only
julia> @time for i in 1:100000; b=bb; end
0.000216 seconds
# You can work with mixed type tuples,
# they are efficient as given
julia> @time for i in 1:100000; b=(1,2.0); end
0.000193 seconds