Round all decimal points in Golang

Question

I'm trying to unconventionally round all the digits in a float64 variable. For example:

3.4444445 --> 3.5

I want to do this without converting it into a string!

3.4444444 rounded to a single decimal point is 3.4, not 3.5. Why do you want code that breaks math? What does 3.644445 "round" to? 3.7? 3.6? 3.65? If what you're doing makes sense, you need to describe it in proper detail, and don't call it "rounding", since obviously that's not what it is. — Jonathan Hall, Aug 28 '18 at 06:58

score 13 · Answer 1 · edited Aug 12 '20 at 12:59

13

Golang's math library provides a Round function. However, it rounds the float64 to an int, meaning decimals are lost.

A quick workaround around this would be to multiple the number by the number of decimals you want to save, then round, then divide it back again:

raw := 4.335
rounded := math.Round(raw * 10) / 10

Will give you the desired result.

You may want to create a little helper function to round saving any number of digits:

func roundTo(n float64, decimals uint32) float64 {
  return math.Round(n*math.Pow(10, float64(decimals))) / math.Pow(10, float64(decimals))
}

Usage:

roundTo(4.2655, 1) // 4.3
roundTo(4.3454, 3) // 4.345

edited Aug 12 '20 at 12:59

Lee

answered Aug 28 '18 at 00:18

user8774937

I think OP wants to round to the nearest half. (or may be not) – zerkms Aug 28 '18 at 00:22
@zerkms Yes. That's what the code does. – user8774937 Aug 28 '18 at 00:25
Would it be `3.4444445 --> 3.5` as OP posted? I'm not sure I can run your code on play: `invalid operation: n * decimals (mismatched types float64 and uint32)` – zerkms Aug 28 '18 at 00:26
There is a bug in the helper function. float64(decimals) should be math.Pow(10, float64(decimals)) – stoffen Aug 05 '20 at 07:58
1

In a situation where you don't know the number of decimal points this works math.Ceil(n*100)/100 100 can be 1000 or the number of your desired decimals. – Marvin Collins May 31 '22 at 14:40

1 Answers1