github.com/asmsh/uniquerand

uniquerand

Provides functionality to generate unique random numbers within a specified range.

Features

• Generate unique random numbers within a specified range.
• Configure the range and the source of random numbers.
• Optimized for short ranges.
• Verify if a number is currently consumed from the specified range.
• Offers a way to reuse (re-generate) a previously generated number.
• Get the number of unique random numbers used from the range so far.
• Modular and easy-to-use API.

Getting Started

Prerequisites

• go1.18 or higher.

Installation

go get github.com/asmsh/uniquerand

Examples

package main

import (
	"fmt"

	"github.com/asmsh/uniquerand"
)

func main() {
	uri := uniquerand.Int{}
	uri.Reset(20)

	for i, ok := uri.Get(); ok; i, ok = uri.Get() {
		// do something with 'i'...
		fmt.Println("i", i)

		// if you are interested in getting the same 'i' value later,
		// return it to the rand source, and it will be randomly returned
		// via a future call to Get.
		// uri.Put(i)
	}

	// print some statistics
	fmt.Println("Num generated:", uri.Count(), "out of", uri.Range())
}

Performance

Some benchmarks (*):

Benchmark_Int/default/Get
Benchmark_Int/default/Get-8     	        53634074	        21.36 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_32/Get-8              	58472168	        21.05 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_64/Get-8              	30796538	        39.14 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_256/Get-8             	 9452564	        124.9 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_1024/Get-8            	 2636010	        454.4 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_4096/Get-8            	  972870	        1780 ns/op	       0 B/op	       0 allocs/op

Benchmark_Int/default/Get_&_Put
Benchmark_Int/default/Get_&_Put-8         	99708832	        12.19 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_32/Get_&_Put-8        	98668995	        11.93 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_64/Get_&_Put-8        	84568580	        14.12 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_256/Get_&_Put-8       	100000000	        11.91 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_1024/Get_&_Put-8      	100000000	        11.78 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_4096/Get_&_Put-8      	100000000	        11.81 ns/op	       0 B/op	       0 allocs/op

Benchmark_Int/default/Rest_&_Get
Benchmark_Int/default/Rest_&_Get-8        	98733601	        12.26 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_32/Rest_&_Get-8       	100000000	        11.84 ns/op	       0 B/op	       0 allocs/op
Benchmark_Int/range_64/Rest_&_Get-8       	61138449	        18.84 ns/op	       4 B/op	       1 allocs/op
Benchmark_Int/range_256/Rest_&_Get-8      	52939521	        22.48 ns/op	      32 B/op	       1 allocs/op
Benchmark_Int/range_1024/Rest_&_Get-8     	41960976	        27.93 ns/op	     128 B/op	       1 allocs/op
Benchmark_Int/range_4096/Rest_&_Get-8     	16797096	        64.83 ns/op	     512 B/op	       1 allocs/op

(*): Benchmarks were done on an M2 Macbook Air.

Theory:

It depends on another Random Number Generator (RNG) (the randomness source) for generating numbers at first.
It checks for the uniqueness of the generated number against a bits-memory.
If the generated number (by the RNG) is found to be not unique, it replaces it with the nearest unique number not used within the specified range.

Limitation: The algorithm for returning the nearest unique number (replacing algo) is not random, however, it could cause the Get method to produce sequential numbers only in the case of the RNG returning the same number over and over.