可以将文章内容翻译成中文,广告屏蔽插件可能会导致该功能失效(如失效,请关闭广告屏蔽插件后再试):
问题:
I am looking for a way to generate all possible permutations of a list of elements. Something similar to python's itertools.permutations(arr)
permutations ([])
[]
permutations ([1])
[1]
permutations ([1,2])
[1, 2]
[2, 1]
permutations ([1,2,3])
[1, 2, 3]
[1, 3, 2]
[2, 1, 3]
[2, 3, 1]
[3, 1, 2]
[3, 2, 1]
With the difference that I do not care whether permutations would be generated on demand (like a generator in python) or all together. I also do not care whether they will be lexicographically sorted. All I need is to somehow get these n!
permutations.
回答1:
There are a lot of the algorithms that generate permutations. One of the easiest I found is Heap's algorithm:
It generates each permutation from the previous one by choosing a pair
of elements to interchange.
The idea and a pseudocode that prints the permutations one after another is outlined in the above link. Here is my implementation of the algorithm which returns all permutations
func permutations(arr []int)[][]int{
var helper func([]int, int)
res := [][]int{}
helper = func(arr []int, n int){
if n == 1{
tmp := make([]int, len(arr))
copy(tmp, arr)
res = append(res, tmp)
} else {
for i := 0; i < n; i++{
helper(arr, n - 1)
if n % 2 == 1{
tmp := arr[i]
arr[i] = arr[n - 1]
arr[n - 1] = tmp
} else {
tmp := arr[0]
arr[0] = arr[n - 1]
arr[n - 1] = tmp
}
}
}
}
helper(arr, len(arr))
return res
}
and here is an example of how to use it (Go playground):
arr := []int{1, 2, 3}
fmt.Println(permutations(arr))
[[1 2 3] [2 1 3] [3 2 1] [2 3 1] [3 1 2] [1 3 2]]
One thing to notice that the permutations are not sorted lexicographically (as you have seen in itertools.permutations
). If for some reason you need it to be sorted, one way I have found it is to generate them from a factorial number system (it is described in permutation section
and allows to quickly find n-th lexicographical permutation).
P.S. you can also take a look at others people code here and here
回答2:
Here's code that iterates over all permutations without generating them all first. The slice p
keeps the intermediate state as offsets in a Fisher-Yates shuffle algorithm. This has the nice property that the zero value for p
describes the identity permutation.
package main
import "fmt"
func nextPerm(p []int) {
for i := len(p) - 1; i >= 0; i-- {
if i == 0 || p[i] < len(p)-i-1 {
p[i]++
return
}
p[i] = 0
}
}
func getPerm(orig, p []int) []int {
result := append([]int{}, orig...)
for i, v := range p {
result[i], result[i+v] = result[i+v], result[i]
}
return result
}
func main() {
orig := []int{11, 22, 33}
for p := make([]int, len(orig)); p[0] < len(p); nextPerm(p) {
fmt.Println(getPerm(orig, p))
}
}
回答3:
don't use recursion! If you want to take advantages of go's concurrency and scalability try something like QuickPerm
package main
import (
"fmt"
)
func main() {
for perm := range GeneratePermutations([]int{1,2,3,4,5,6,7}){
fmt.Println(perm)
}
}
func GeneratePermutations(data []int) <-chan []int {
c := make(chan []int)
go func(c chan []int) {
defer close(c)
permutate(c, data)
}(c)
return c
}
func permutate(c chan []int, inputs []int){
output := make([]int, len(inputs))
copy(output, inputs)
c <- output
size := len(inputs)
p := make([]int, size + 1)
for i := 0; i < size + 1; i++ {
p[i] = i;
}
for i := 1; i < size;{
p[i]--
j := 0
if i % 2 == 1 {
j = p[i]
}
tmp := inputs[j]
inputs[j] = inputs[i]
inputs[i] = tmp
output := make([]int, len(inputs))
copy(output, inputs)
c <- output
for i = 1; p[i] == 0; i++{
p[i] = i
}
}
}
回答4:
In my case I had a reference to an array, then I've did a few changes in your example:
func generateIntPermutations(array []int, n int, result *[][]int) {
if n == 1 {
dst := make([]int, len(array))
copy(dst, array[:])
*result = append(*result, dst)
} else {
for i := 0; i < n; i++ {
generateIntPermutations(array, n-1, result)
if n%2 == 0 {
// Golang allow us to do multiple assignments
array[0], array[n-1] = array[n-1], array[0]
} else {
array[i], array[n-1] = array[n-1], array[i]
}
}
}
}
numbers := []int{0, 1, 2}
var result [][]int
generateIntPermutations(numbers, len(numbers), &result)
// result -> [[0 1 2] [1 0 2] [2 1 0] [1 2 0] [2 0 1] [0 2 1]]