All the integer functions in math/rand generate non-negative numbers.
rand.Int() int // [0, MaxInt]
rand.Int31() int32 // [0, MaxInt32]
As to prevent repeating min
and max
over and over again, I suggest to switch range and random in thinking about it. This is what I found to work as expected:
package main
import (
"fmt"
"math/rand"
)
// range specification, note that min <= max
type IntRange struct {
min, max int
}
// get next random value within the interval including min and max
func (ir *IntRange) NextRandom(r* rand.Rand) int {
return r.Intn(ir.max - ir.min +1) + ir.min
}
func main() {
r := rand.New(rand.NewSource(55))
ir := IntRange{-1,1}
for i := 0; i<10; i++ {
fmt.Println(ir.NextRandom(r))
}
}
See on Go Playground
The solution you found in the Cookbook misses to exactly specify how min
and max
work, but of course it meets your specification ([-min, max)). I decided to specify the range as a closed interval ([-min, max], that means its borders are included in the valid range). Compared to my understanding of the Cookbook description:
gives you that random number within any two positive numbers that you specify (in this case, 1 and 6).
(which can be found below the code snippet in the Golang Cookbook)
the Cookbook implementation is off by one (which of course brings it in good company with lots of programs that are helpful at first glance).