On Nov 26, 2015 9:15 PM, "Sheehan Olver" <[email protected]> wrote: > > OK so it's a bit inconsistent then with BigFloat: > > julia> BigFloat(1/3)==(1/3) > > true
You are converting a Float64 1/3 to a big float and of course they are equal. One way to get a higher precision, use sth like big(1) / 3 > > > julia> BigFloat(1/3)==(1//3) > > false > > > > > On Friday, November 27, 2015 at 12:40:12 PM UTC+11, Yichao Yu wrote: >> >> >> On Nov 26, 2015 8:27 PM, "Sheehan Olver" <[email protected]> wrote: >> > >> > >> > >> > This really surprised me: >> > >> > julia> (1.0π) ≠ π >> > >> > true >> > >> > >> > >> > I guess π is treated as infinite precision until it becomes a float? >> >> Right.
