[julia-users] How does Julia decide where to truncate displayed values in the REPL?

[gael . mcdon]

Hi,

I was experimenting with @simd and was a bit surprised about some results 
on different implementations of a plain summing function:

julia> f(a)
399921.25f0

julia> g(a)
399916.2f0

julia> sum(a)
399922.25f0

julia> sum_kbn(a)
399920.66f0

julia> @printf("%.6f",g(a))
399916.187500

I don't understand the behavior here: most of the time (not to say always 
when using the REPL), I got 8 digits printed with 32 bit floats. The only 
exception is when there are less than 8 significant digits. But here it's 
not the case: why isn't g(a) displayed with 8 digits?

As an additional question: is there advanced string formatting available in 
Julia? Like a format function or a judicious rem overloading? 

Cheers,
Gaël







a = rand(Float32,800000)

function f{T}(a::Array{T,1})
   s = zero(T)
   @simd for i in 1:length(a)
       @inbounds s = s + a[i]
   end
   return s
end

function g{T}(a::Array{T,1})
   s = zero(T)
   for i in 1:length(a)
       @inbounds s = s + a[i]
   end
   return s
end