[Go] The Error of Floating Point Number

Like most programming languages, Go also has the floating point error problem.

Problem

Please look at the below code.

package main

import (
  "fmt"
)

func main() {
  a := 0.1
  b := 0.2
  c := a + b
  fmt.Println(c)
}
// 0.30000000000000004

c was not a 0.3!

Therefore, we are hard to compare floating point numbers.

a := 0.1
b := 0.2
c := 0.3
d := a + b
if c == d {
  fmt.Println("It's the same!")
} else {
  fmt.Println("It's different!")
}
// It's different!

Them, how can we achieve this?

There are two possible solutions.

Solution1: math.Nextafter()

math package has Nextafter() function.

func Nextafter(x, y float64) (r float64)

Nextafter() function returns the next representable float64 value after x towards y.

if math.Nextafter(c, d) == d {
  fmt.Println("It's the same!")
} else {
  fmt.Println("It's different!")
}
// It's the same!

Solution2: math/big

math package has the big package and it can handle floating point numbers precisely.

a, _ := new(big.Float).SetString("0.1")
b, _ := new(big.Float).SetString("0.2")
c, _ := new(big.Float).SetString("0.3")
d := new(big.Float).Add(a, b)
if c.Cmp(d) == 0 {
  fmt.Println("It's the same!")
} else {
  fmt.Println("It's different!")
}
// It's the same!

It's a little complex but if you handle a financial program, you'd like to use this package.