Idiomatic/Proper Go code refactoring for a UnionFind library - go

I am practicing a classic algorithm problem in go "number of islands". I want to solve it using unionfind. Although I can tweak and make it work, I would like to know the best way to structure my code.
Here is the main program.
package main
import (
"fmt"
u "practice/leetcode/library/unionfind"
)
type point u.Point
func numIslands(grid [][]byte) int {
res := 0
if grid == nil || grid[0] == nil {
return res
}
m := len(grid)
n := len(grid[0])
num := 0
uf := u.NewUnionFind()
directions := []point{
point{1, 0},
point{-1, 0},
point{0, 1},
point{0, -1},
}
emptyPoint := point{}
for i := 0; i < m; i++ {
for j := 0; j < n; j++ {
if grid[i][j] == 1 {
p := point{i, j}
uf.Add(p)
num++
for _, v := range directions {
newx := i + v.X
newy := j + v.Y
if 0 <= newx && newx < m && 0 <= newy && newy < n && grid[newx][newy] == 1 {
newp := point{newx, newy}
if uf.Find(newp) == emptyPoint {
continue
}
uf.Union(p, newp)
num--
}
}
}
}
}
return num
}
func main() {
// expect 1
grid := [][]byte {
{1,1,1,1,0},
{1,1,0,1,0},
{1,1,0,0,0},
{0,0,0,0,0},
}
fmt.Println(numIslands(grid))
// expect 2
grid = [][]byte {
{1,0},
{0,1},
}
fmt.Println(numIslands(grid))
}
Here is the unionfind library I wrote
package unionfind
type Point struct {
X int
Y int
}
type UnionFind struct {
parent map[Point]Point
}
func NewUnionFind() *UnionFind {
parent := make(map[Point]Point)
return &UnionFind{parent}
}
func (uf *UnionFind) Add(c Point) {
if _, ok := uf.parent[c]; ok {
return
}
uf.parent[c] = c
}
func (uf *UnionFind) Find(c Point) Point {
if p, ok := uf.parent[c]; ok {
if p != c {
uf.parent[c] = uf.Find(p)
}
return uf.parent[c]
}
return Point{}
}
func (uf *UnionFind) Union(c1 Point, c2 Point) {
p1 := uf.Find(c1)
p2 := uf.Find(c2)
if p1 != p2 {
uf.parent[p1] = p2
}
}
The problem is when I ran the main program, i got the following errors:
# command-line-arguments
./200_number_of_islands_uf.go:31:15: cannot use p (type point) as type unionfind.Point in argument to uf.Add
./200_number_of_islands_uf.go:38:23: cannot use newp (type point) as type unionfind.Point in argument to uf.Find
./200_number_of_islands_uf.go:38:30: invalid operation: uf.Find(newp) == emptyPoint (mismatched types unionfind.Point and point)
./200_number_of_islands_uf.go:41:21: cannot use p (type point) as type unionfind.Point in argument to uf.Union
./200_number_of_islands_uf.go:41:21: cannot use newp (type point) as type unionfind.Point in argument to uf.Union
The hard requirement I want is:
1.I want to keep the unionfind as a library
2.I need to access the Point struct in the main program
The design requirement I am trying to follow is:
I tried to avoid using an interface for the UnionFind parent map since I can foresee only the Point struct get pass in to the library.
My requirements could be wrong. I am open to suggestions to refactor the code and make it more elegant looking.

When you do this:
type point u.Point
then point is not just an aliased name for u.Point which can be used interchangeably - it's an entirely new type, and one which your uniontype package knows nothing about and therefore will not accept.
So just don't do that, and instead directly use the type the uniontype package provides for you. For example, change:
directions := []point{
point{1, 0},
point{-1, 0},
point{0, 1},
point{0, -1},
}
emptyPoint := point{}
to:
directions := []u.Point{
u.Point{1, 0},
u.Point{-1, 0},
u.Point{0, 1},
u.Point{0, -1},
}
emptyPoint := u.Point{}
and so on.

Related

Golang reverse a arbitrary slice

var Reverse = func(slice interface{}) {
s := reflect.ValueOf(slice)
// if s is a pointer of slice
if s.Kind() == reflect.Ptr {
s = s.Elem()
}
i := 0
j := s.Len() - 1
for i < j {
x, y := s.Index(i).Interface(),
s.Index(j).Interface()
s.Index(i).Set(reflect.ValueOf(y))
s.Index(j).Set(reflect.ValueOf(x))
i++
j--
}
}
I found this way works.But it is not elegant....
I know there is a method "Swapper" in reflect package. But I don't know how to make it work if the argument of above function is a pointer of slice.
Really appreciate.
Your code works fine. To use reflect.Swapper, just pass s.Interface() into it:
var Reverse = func(slice interface{}) {
s := reflect.ValueOf(slice)
// if s is a pointer of slice
if s.Kind() == reflect.Ptr {
s = s.Elem()
}
swp := reflect.Swapper(s.Interface())
for i,j :=0,s.Len() - 1; i<j; i,j = i+1,j-1 {
swp(i,j)
}
}
Playground: https://play.golang.org/p/DSq_iZRZX4b

Change slice content and capacity inside a function in-place

I am trying to learn Go, so here is my very simple function for removing adjacent duplicates from slice for exercise from the book by Donovan & Kernighan.
Here is the code: https://play.golang.org/p/avHc1ixfck
package main
import "fmt"
func main() {
a := []int{0, 1, 1, 3, 3, 3}
removeDup(a)
fmt.Println(a)
}
func removeDup(s []int) {
n := len(s)
tmp := make([]int, 0, n)
tmp = append(tmp, s[0])
j := 1
for i := 1; i < n; i++ {
if s[i] != s[i-1] {
tmp = append(tmp, s[i])
j++
}
}
s = s[:len(tmp)]
copy(s, tmp)
}
It should print out [0 1 3] - and I checked, actually tmp at the end of the function it has desired form. However, the result is [0 1 3 3 3 3]. I guess there is something with copy function.
Can I somehow replace input slice s with the temp or trim it to desired length?
Option 1
Return a new slice as suggested by #zerkms.
https://play.golang.org/p/uGJiD3WApS
package main
import "fmt"
func main() {
a := []int{0, 1, 1, 3, 3, 3}
a = removeDup(a)
fmt.Println(a)
}
func removeDup(s []int) []int {
n := len(s)
tmp := make([]int, 0, n)
tmp = append(tmp, s[0])
for i := 1; i < n; i++ {
if s[i] != s[i-1] {
tmp = append(tmp, s[i])
}
}
return tmp
}
Option 2
Use pointers for pass-by-reference.
The same thing in effect as that of option1.
https://play.golang.org/p/80bE5Qkuuj
package main
import "fmt"
func main() {
a := []int{0, 1, 1, 3, 3, 3}
removeDup(&a)
fmt.Println(a)
}
func removeDup(sp *[]int) {
s := *sp
n := len(s)
tmp := make([]int, 0, n)
tmp = append(tmp, s[0])
for i := 1; i < n; i++ {
if s[i] != s[i-1] {
tmp = append(tmp, s[i])
}
}
*sp = tmp
}
Also, refer to following SO thread:
Does Go have no real way to shrink a slice? Is that an issue?
Here's two more slightly different ways to achieve what you want using sets and named types. The cool thing about named types is that you can create interfaces around them and can help with the readability of lots of code.
package main
import "fmt"
func main() {
// returning a list
a := []int{0, 1, 1, 3, 3, 3}
clean := removeDup(a)
fmt.Println(clean)
// creating and using a named type
nA := &newArrType{0, 1, 1, 3, 3, 3}
nA.removeDup2()
fmt.Println(nA)
// or... casting your orginal array to the named type
nB := newArrType(a)
nB.removeDup2()
fmt.Println(nB)
}
// using a set
// order is not kept, but a set is returned
func removeDup(s []int) (newArr []int) {
set := make(map[int]struct{})
for _, n := range s {
set[n] = struct{}{}
}
newArr = make([]int, 0, len(set))
for k := range set {
newArr = append(newArr, k)
}
return
}
// using named a typed
type newArrType []int
func (a *newArrType) removeDup2() {
x := *a
for i := range x {
f := i + 1
if f < len(x) {
if x[i] == x[f] {
x = x[:f+copy(x[f:], x[f+1:])]
}
}
}
// check the last 2 indexes
if x[len(x)-2] == x[len(x)-1] {
x = x[:len(x)-1+copy(x[len(x)-1:], x[len(x)-1+1:])]
}
*a = x
}

Indirectly change a value in a struct in Go

I have the following code, feel free to offer pointers if you wish:
package main
import (
"fmt"
)
type Grid struct {
rows int
cols int
tiles []Tile
}
type Tile struct {
x int
y int
contents int
}
func (g Grid) AddTile(t Tile) {
g.tiles = append(g.tiles, t)
}
func (g *Grid) Row(num int) []Tile {
numTiles := len(g.tiles)
row := []Tile{}
for i := 0; i < numTiles; i++ {
tile := g.tiles[i]
if (tile.y == num) {
row = append(row, tile)
}
}
return row
}
/*
HERE IS WHERE I NEED HELP
*/
func (g *Grid) SetRow(num, val int) {
row := g.Row(num)
rowLength := len(row)
for i := 0; i < rowLength; i++ {
tile := &row[i]
tile.contents = val
}
}
func (g Grid) Col(num int) []Tile {
numTiles := len(g.tiles)
col := []Tile{}
for i := 0; i < numTiles; i++ {
tile := g.tiles[i]
if (tile.x == num) {
col = append(col, tile)
}
}
return col
}
func MakeTile(x, y int) Tile {
tile := Tile{x: x, y: y}
return tile
}
func MakeGrid(rows, cols int) Grid {
g := Grid{ rows: rows, cols: cols}
for r := 1; r <= rows; r++ {
for c := 1; c <= cols; c++ {
g.tiles = append(g.tiles, MakeTile(r, c))
}
}
return g
}
func main() {
g := MakeGrid(256, 256)
g.SetRow(100, 5)
fmt.Println(g.Row(100))
}
I am doing this, more than anything, as a simple project to help me learn Go. The problem that is have run in to is here
/*
HERE IS WHERE I NEED HELP
*/
func (g *Grid) SetRow(num, val int) {
row := g.Row(num)
rowLength := len(row)
for i := 0; i < rowLength; i++ {
tile := &row[i]
tile.contents = val
}
}
Somewhere it seems like I need to be making a pointer to the actual Tiles that I'm trying to modify. As it is the SetRow function doesn't actually modify anything. What am I doing wrong? Keep in mind I just started learning Go 2 days ago, so this is a learning experience :)
One way to accomplish your goal is to use pointers to tiles throughout the code. Change the Grid tiles field to:
tiles []*Tile
and several related changes through the code.
Also, change all the methods to use pointer receivers. The AddTile method as written in the question discards the modification to the grid on return.
playground example

Is this a reasonable and idiomatic GoLang circular shift implementation?

Can anyone comment on whether this is a reasonable and idiomatic way of implementing circular shift of integer arrays in Go? (I deliberately chose not to use bitwise operations.)
How could it be improved?
package main
import "fmt"
func main() {
a := []int{1,2,3,4,5,6,7,8,9,10}
fmt.Println(a)
rotateR(a, 5)
fmt.Println(a)
rotateL(a, 5)
fmt.Println(a)
}
func rotateL(a []int, i int) {
for count := 1; count <= i; count++ {
tmp := a[0]
for n := 1;n < len(a);n++ {
a[n-1] = a[n]
}
a[len(a)-1] = tmp
}
}
func rotateR(a []int, i int) {
for count := 1; count <= i; count++ {
tmp := a[len(a)-1]
for n := len(a)-2;n >=0 ;n-- {
a[n+1] = a[n]
}
a[0] = tmp
}
}
Rotating the slice one position at a time, and repeating to get the total desired rotation means it will take time proportional to rotation distance × length of slice. By moving each element directly into its final position you can do this in time proportional to just the length of the slice.
The code for this is a little more tricky than you have, and you’ll need a GCD function to determine how many times to go through the slice:
func gcd(a, b int) int {
for b != 0 {
a, b = b, a % b
}
return a
}
func rotateL(a []int, i int) {
// Ensure the shift amount is less than the length of the array,
// and that it is positive.
i = i % len(a)
if i < 0 {
i += len(a)
}
for c := 0; c < gcd(i, len(a)); c++ {
t := a[c]
j := c
for {
k := j + i
// loop around if we go past the end of the slice
if k >= len(a) {
k -= len(a)
}
// end when we get to where we started
if k == c {
break
}
// move the element directly into its final position
a[j] = a[k]
j = k
}
a[j] = t
}
}
Rotating a slice of size l right by p positions is equivalent to rotating it left by l − p positions, so you can simplify your rotateR function by using rotateL:
func rotateR(a []int, i int) {
rotateL(a, len(a) - i)
}
Your code is fine for in-place modification.
Don't clearly understand what you mean by bitwise operations. Maybe this
package main
import "fmt"
func main() {
a := []int{1, 2, 3, 4, 5, 6, 7, 8, 9, 10}
fmt.Println(a)
rotateR(&a, 4)
fmt.Println(a)
rotateL(&a, 4)
fmt.Println(a)
}
func rotateL(a *[]int, i int) {
x, b := (*a)[:i], (*a)[i:]
*a = append(b, x...)
}
func rotateR(a *[]int, i int) {
x, b := (*a)[:(len(*a)-i)], (*a)[(len(*a)-i):]
*a = append(b, x...)
}
Code works https://play.golang.org/p/0VtiRFQVl7
It's called reslicing in Go vocabulary. Tradeoff is coping and looping in your snippet vs dynamic allocation in this. It's your choice, but in case of shifting 10000 elements array by one position reslicing looks much cheaper.
I like Uvelichitel solution but if you would like modular arithmetic which would be O(n) complexity
package main
func main(){
s := []string{"1", "2", "3"}
rot := 5
fmt.Println("Before RotL", s)
fmt.Println("After RotL", rotL(rot, s))
fmt.Println("Before RotR", s)
fmt.Println("After RotR", rotR(rot,s))
}
func rotL(m int, arr []string) []string{
newArr := make([]string, len(arr))
for i, k := range arr{
newPos := (((i - m) % len(arr)) + len(arr)) % len(arr)
newArr[newPos] = k
}
return newArr
}
func rotR(m int, arr []string) []string{
newArr := make([]string, len(arr))
for i, k := range arr{
newPos := (i + m) % len(arr)
newArr[newPos] = k
}
return newArr
}
If you need to enter multiple values, whatever you want (upd code Uvelichitel)
package main
import "fmt"
func main() {
var N, n int
fmt.Scan(&N)
a := make([]int, N)
for i := 0; i < N; i++ {
fmt.Scan(&a[i])
}
fmt.Scan(&n)
if n > 0 {
rotateR(&a, n%len(a))
} else {
rotateL(&a, (n*-1)%len(a))
}
for _, elem := range a {
fmt.Print(elem, " ")
}
}
func rotateL(a *[]int, i int) {
x, b := (*a)[:i], (*a)[i:]
*a = append(b, x...)
}
func rotateR(a *[]int, i int) {
x, b := (*a)[:(len(*a)-i)], (*a)[(len(*a)-i):]
*a = append(b, x...)
}

What's wrong with my Priority Queue's Pop method?

Based on Rob Pike's load balancer demo, I implemented my own priority queue, but my Pop method is not right, can anyone tell me what's wrong?
package main
import (
"fmt"
"container/heap"
)
type ClassRecord struct {
name string
grade int
}
type RecordHeap []*ClassRecord
func (p RecordHeap) Len() int { return len(p) }
func (p RecordHeap) Less(i, j int) bool {
return p[i].grade < p[j].grade
}
func (p *RecordHeap) Swap(i, j int) {
a := *p
a[i], a[j] = a[j], a[i]
}
func (p *RecordHeap) Push(x interface{}) {
a := *p
n := len(a)
a = a[0 : n+1]
r := x.(*ClassRecord)
a[n] = r
*p = a
}
func (p *RecordHeap) Pop() interface{} {
a := *p
*p = a[0 : len(a)-1]
r := a[len(a)-1]
return r
}
func main() {
a := make([]ClassRecord, 6)
a[0] = ClassRecord{"John", 80}
a[1] = ClassRecord{"Dan", 85}
a[2] = ClassRecord{"Aron", 90}
a[3] = ClassRecord{"Mark", 65}
a[4] = ClassRecord{"Rob", 99}
a[5] = ClassRecord{"Brian", 78}
h := make(RecordHeap, 0, 100)
for _, c := range a {
fmt.Println(c)
heap.Push(&h, &c)
fmt.Println("Push: heap has", h.Len(), "items")
}
for i, x := 0, heap.Pop(&h).(*ClassRecord); i < 10 && x != nil; i++ {
fmt.Println("Pop: heap has", h.Len(), "items")
fmt.Println(*x)
}
}
EDIT: besides the way cthom06 pointed out, another way to fix this is to create a pointer array as follows,
a := make([]*ClassRecord, 6)
a[0] = &ClassRecord{"John", 80}
a[1] = &ClassRecord{"Dan", 85}
......
EDIT:
Oh I should've seen this right away.
heap.Push(&h, &c)
You push the address of c, which gets reused on each iteration of range. Every record in the heap is a pointer to the same area in memory, which ends up being Brian. I'm not sure if this is intended behavior or a compiler bug, but
t := c
heap.Push(&h, &t)
works around it.
Also: Your for loop is wrong.
for h.Len() > 0 {
x := heap.Pop(&h...
should fix it.

Resources