I will add that this is why most financial applications using “fixed place” or “integer math” libraries - to have control over these sorts of issues.
An example is robaho/fixed <http://github.com/robaho/fixed> > On Oct 10, 2022, at 6:35 PM, Dante Castagnoli <dante.castagn...@gmail.com> > wrote: > > Thanks, Rob, > > As the behavior is as in "C", the behavior is as it should be. I also > understand that "1" is a closer representation than 0.9999 should the number > be 0.99999, though it's a choice as it diverges from string-type formats. > Regardless of whether the choice was right or wrong, it is immaterial in my > view, as any change would confuse those familiar with the current behavior. > > I do think the documentation I cited above could be clearer, and without > difficulty. The documentation of "precision" leads one to believe that > additional digits will be printed, leaving one with the impression those > digits will be as they are in the float, and not rounded. > > Thanks for the response. My view of the issue hasn't really changed given > the responses. I'm simply considering how this might be easier for others, > as I now understand the behavior. That's all. > > D. > > > > On Mon, Oct 10, 2022 at 4:23 PM Rob Pike <r...@golang.org > <mailto:r...@golang.org>> wrote: > This behavior is how floating-point math works, which is not best addressed > in the documentation for the formatted I/O library. We see may comments from > people new to floating point who argue that there are bugs in Go's > arithmetic, but it's invariably a lack of familiarity with the—admittedly > rather tricky—properties of imprecise arithmetic. > > -rob > > > On Tue, Oct 11, 2022 at 9:01 AM Dante Castagnoli <dante.castagn...@gmail.com > <mailto:dante.castagn...@gmail.com>> wrote: > Thanks! > > It's not lost, though. It shows up with the %g form. > > Knowing about %g, my issue is not extant, but others might hit it. I'm > relatively new to floating point issues, and I saw numbers showing up as > integers, and it took a while to sort out what was going on, is all. > > As the behavior matches "c", I would suggest at minimum making the > documentation clearer. > > https://go.dev/play/p/WBlf3jltimr <https://go.dev/play/p/WBlf3jltimr> > > On Monday, October 10, 2022 at 10:27:00 AM UTC-7 Brian Candler wrote: > Whilst Go handles constants with arbitrary precision AFAIK, precision is lost > once values are assigned to a float64 variable (or passed as an argument to a > function). There are only 53 bits of precision in the "mantissa". > > For more info, you might find this helpful: > https://0.30000000000000004.com/ <https://0.30000000000000004.com/> > > On Monday, 10 October 2022 at 17:48:56 UTC+1 dante.ca...@gmail.com <> wrote: > Hi, > > I'm wondering what folks think about the behavior of %f and precision. > > The go fmt documentation states: > > For floating-point values, width sets the minimum width of the field and > > precision sets the number of places after the decimal, > > What I discovered is there is rounding going on in %f as shown below. > > https://go.dev/play/p/23AfRtwOqcD <https://go.dev/play/p/23AfRtwOqcD> > > I found this behavior as I'm converting floats to integers, and was testing > fine regions between digits, and experienced the behavior. I didn't expect > it. > > I expected the behavior of printing "n" characters of a string. However, I > note that "c" behaves the same way: > > $ cat math.c > #include<stdio.h> > int main(int argc, char *argv[]) { > printf("%0.3f\n", 0.9999999999); > } > $ ./math > 1.000 > $ > > -- > You received this message because you are subscribed to the Google Groups > "golang-nuts" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to golang-nuts+unsubscr...@googlegroups.com > <mailto:golang-nuts+unsubscr...@googlegroups.com>. > To view this discussion on the web visit > https://groups.google.com/d/msgid/golang-nuts/4a15ac2c-e30d-4d6e-be8f-251996fb01cen%40googlegroups.com > > <https://groups.google.com/d/msgid/golang-nuts/4a15ac2c-e30d-4d6e-be8f-251996fb01cen%40googlegroups.com?utm_medium=email&utm_source=footer>. > > -- > You received this message because you are subscribed to the Google Groups > "golang-nuts" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to golang-nuts+unsubscr...@googlegroups.com > <mailto:golang-nuts+unsubscr...@googlegroups.com>. > To view this discussion on the web visit > https://groups.google.com/d/msgid/golang-nuts/CAEaSmXcfd3MO3eRjHkshi%3DMg3G9RrN9rH1B%3DSoJVth3BDG5-dg%40mail.gmail.com > > <https://groups.google.com/d/msgid/golang-nuts/CAEaSmXcfd3MO3eRjHkshi%3DMg3G9RrN9rH1B%3DSoJVth3BDG5-dg%40mail.gmail.com?utm_medium=email&utm_source=footer>. -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/61447183-72B3-4568-876F-1944FCA11F32%40ix.netcom.com.
