Related
I would like to create a nested loop where the nesting depth determined by the size on an array. For example I have an array of integers and I would like to check all tuples where . Is there a way to do this simply in c# (or any other language)?
The only easy idea I had was to just multiply the numbers and do a for loop up to that number but unfortunately the product reaches the int limit. Also in this case I have no way to add extra conditions on each level.
Here's a small example and the product approach:
int[] a = new int[]{2, 3, 2}; //we have an array like this. In the description above it is a_1, a_2, a_3
void f(int[] i) //I have a function that works on an array input
{...}
bool check(int[] i) //I have a checker function, with an array parameter too
{...}
//And I would like to perform this function on all these arrays
//f({0, 0, 0})
//f({0, 0, 1})
//010
//011
//020
//021
//100
//101
//110
//111
//120
//f({1, 2, 1})
//But if let's say check({1, 0}) is false then don't perform the f function on {1, 0, 0} and {1, 0, 1}
//A simple implementation if we know the size of the array a is the following
for(int i1=0;i1<a[0];i1++)
{
if(!check({i1}))
continue;
for(int i2=0;i2<a[1];i2++)
{
if(!check({i1, i2}))
continue;
for(int i3=0;i3<a[2];i3++)
{
if(!check({i1, i2, i3}))
continue;
f({i1, i2, i3});
}
}
}
//But this obviously fails as we have no idea apriori of the size of the array a
//An alternative I have is the following:
int prod = 1;
foreach(int x in a)
{
prod *= x;
}
for(int c=0;c<prod;c++)
{
int d=c;
int[] i = new int[a.Length];
for(int l=0;l<a.Length;l++)
{
i[l]=d%(a[l]);
d /= a[l];
}
f(i);
}
//But the problem with this implementation is that in my case prod is larger than the int limit. Also this loops through all the incorrect cases too, where the check function can highly reduce the number of cases to calculate.
I've managed to solve the problem. The idea is that increasing the i value by one is really easy and we can check the condition that we don't overstep the a values and don't violate the check function easily. Below is a code that works
int[] a = ...;
void f(int[] i){...}
bool check(int[] i){...}
int n = a.Length;
int[] i = new int[n];
while (true) //keep increasing the coordinates of i, while we can
{
for (int l = 0; l < n; l++)
{
int[] il = copyFirst(i, l);
while(!check(il)) //check for all first few coordinates if it is correct, skip if incorrect.
//There is a way to improve this even further, as we don't have to check the first few
// coords if it was correct before, so should only care about the recently changed section
{
i = increase(a, i, l);
if (i == null)
{
return;
}
else
{
il = copyFirst(i, l);
}
}
}
f(i);
i = increase(a, i, n-1);
if (i == null) return;
}
int[] copyFirst(int[] i, int l) //this is just a small helper function to copy the first few elements of i, to use in the check
{
int[] ret = new int[l];
for (int k = 0; k < l; k++)
{
ret[k] = i[k];
}
return ret;
}
int[] increase(int[] a, int[] i, int l) //this results in the increased vector and in null if we've reached the end
{
for (int k = l; k >= 0; k--)
{
i[k] = i[k] + 1;
if (i[k] >= a[k])
{
i[k] = 0;
}
else
{
return i;
}
}
return null;
}
I have an array of Foo objects. How do I remove the second element of the array?
I need something similar to RemoveAt() but for a regular array.
If you don't want to use List:
var foos = new List<Foo>(array);
foos.RemoveAt(index);
return foos.ToArray();
You could try this extension method that I haven't actually tested:
public static T[] RemoveAt<T>(this T[] source, int index)
{
T[] dest = new T[source.Length - 1];
if( index > 0 )
Array.Copy(source, 0, dest, 0, index);
if( index < source.Length - 1 )
Array.Copy(source, index + 1, dest, index, source.Length - index - 1);
return dest;
}
And use it like:
Foo[] bar = GetFoos();
bar = bar.RemoveAt(2);
The nature of arrays is that their length is immutable. You can't add or delete any of the array items.
You will have to create a new array that is one element shorter and copy the old items to the new array, excluding the element you want to delete.
So it is probably better to use a List instead of an array.
I use this method for removing an element from an object array. In my situation, my arrays are small in length. So if you have large arrays you may need another solution.
private int[] RemoveIndices(int[] IndicesArray, int RemoveAt)
{
int[] newIndicesArray = new int[IndicesArray.Length - 1];
int i = 0;
int j = 0;
while (i < IndicesArray.Length)
{
if (i != RemoveAt)
{
newIndicesArray[j] = IndicesArray[i];
j++;
}
i++;
}
return newIndicesArray;
}
LINQ one-line solution:
myArray = myArray.Where((source, index) => index != 1).ToArray();
The 1 in that example is the index of the element to remove -- in this example, per the original question, the 2nd element (with 1 being the second element in C# zero-based array indexing).
A more complete example:
string[] myArray = { "a", "b", "c", "d", "e" };
int indexToRemove = 1;
myArray = myArray.Where((source, index) => index != indexToRemove).ToArray();
After running that snippet, the value of myArray will be { "a", "c", "d", "e" }.
This is a way to delete an array element, as of .Net 3.5, without copying to another array - using the same array instance with Array.Resize<T>:
public static void RemoveAt<T>(ref T[] arr, int index)
{
for (int a = index; a < arr.Length - 1; a++)
{
// moving elements downwards, to fill the gap at [index]
arr[a] = arr[a + 1];
}
// finally, let's decrement Array's size by one
Array.Resize(ref arr, arr.Length - 1);
}
Here is an old version I have that works on version 1.0 of the .NET framework and does not need generic types.
public static Array RemoveAt(Array source, int index)
{
if (source == null)
throw new ArgumentNullException("source");
if (0 > index || index >= source.Length)
throw new ArgumentOutOfRangeException("index", index, "index is outside the bounds of source array");
Array dest = Array.CreateInstance(source.GetType().GetElementType(), source.Length - 1);
Array.Copy(source, 0, dest, 0, index);
Array.Copy(source, index + 1, dest, index, source.Length - index - 1);
return dest;
}
This is used like this:
class Program
{
static void Main(string[] args)
{
string[] x = new string[20];
for (int i = 0; i < x.Length; i++)
x[i] = (i+1).ToString();
string[] y = (string[])MyArrayFunctions.RemoveAt(x, 3);
for (int i = 0; i < y.Length; i++)
Console.WriteLine(y[i]);
}
}
Not exactly the way to go about this, but if the situation is trivial and you value your time, you can try this for nullable types.
Foos[index] = null
and later check for null entries in your logic..
Try below code:
myArray = myArray.Where(s => (myArray.IndexOf(s) != indexValue)).ToArray();
or
myArray = myArray.Where(s => (s != "not_this")).ToArray();
As usual, I'm late to the party...
I'd like to add another option to the nice solutions list already present. =)
I would see this as a good opportunity for Extensions.
Reference:
http://msdn.microsoft.com/en-us/library/bb311042.aspx
So, we define some static class and in it, our Method.
After that, we can use our extended method willy-nilly. =)
using System;
namespace FunctionTesting {
// The class doesn't matter, as long as it's static
public static class SomeRandomClassWhoseNameDoesntMatter {
// Here's the actual method that extends arrays
public static T[] RemoveAt<T>( this T[] oArray, int idx ) {
T[] nArray = new T[oArray.Length - 1];
for( int i = 0; i < nArray.Length; ++i ) {
nArray[i] = ( i < idx ) ? oArray[i] : oArray[i + 1];
}
return nArray;
}
}
// Sample usage...
class Program {
static void Main( string[] args ) {
string[] myStrArray = { "Zero", "One", "Two", "Three" };
Console.WriteLine( String.Join( " ", myStrArray ) );
myStrArray = myStrArray.RemoveAt( 2 );
Console.WriteLine( String.Join( " ", myStrArray ) );
/* Output
* "Zero One Two Three"
* "Zero One Three"
*/
int[] myIntArray = { 0, 1, 2, 3 };
Console.WriteLine( String.Join( " ", myIntArray ) );
myIntArray = myIntArray.RemoveAt( 2 );
Console.WriteLine( String.Join( " ", myIntArray ) );
/* Output
* "0 1 2 3"
* "0 1 3"
*/
}
}
}
Here's how I did it...
public static ElementDefinitionImpl[] RemoveElementDefAt(
ElementDefinition[] oldList,
int removeIndex
)
{
ElementDefinitionImpl[] newElementDefList = new ElementDefinitionImpl[ oldList.Length - 1 ];
int offset = 0;
for ( int index = 0; index < oldList.Length; index++ )
{
ElementDefinitionImpl elementDef = oldList[ index ] as ElementDefinitionImpl;
if ( index == removeIndex )
{
// This is the one we want to remove, so we won't copy it. But
// every subsequent elementDef will by shifted down by one.
offset = -1;
}
else
{
newElementDefList[ index + offset ] = elementDef;
}
}
return newElementDefList;
}
In a normal array you have to shuffle down all the array entries above 2 and then resize it using the Resize method. You might be better off using an ArrayList.
private int[] removeFromArray(int[] array, int id)
{
int difference = 0, currentValue=0;
//get new Array length
for (int i=0; i<array.Length; i++)
{
if (array[i]==id)
{
difference += 1;
}
}
//create new array
int[] newArray = new int[array.Length-difference];
for (int i = 0; i < array.Length; i++ )
{
if (array[i] != id)
{
newArray[currentValue] = array[i];
currentValue += 1;
}
}
return newArray;
}
Here's a small collection of helper methods I produced based on some of the existing answers. It utilizes both extensions and static methods with reference parameters for maximum idealness:
public static class Arr
{
public static int IndexOf<TElement>(this TElement[] Source, TElement Element)
{
for (var i = 0; i < Source.Length; i++)
{
if (Source[i].Equals(Element))
return i;
}
return -1;
}
public static TElement[] Add<TElement>(ref TElement[] Source, params TElement[] Elements)
{
var OldLength = Source.Length;
Array.Resize(ref Source, OldLength + Elements.Length);
for (int j = 0, Count = Elements.Length; j < Count; j++)
Source[OldLength + j] = Elements[j];
return Source;
}
public static TElement[] New<TElement>(params TElement[] Elements)
{
return Elements ?? new TElement[0];
}
public static void Remove<TElement>(ref TElement[] Source, params TElement[] Elements)
{
foreach (var i in Elements)
RemoveAt(ref Source, Source.IndexOf(i));
}
public static void RemoveAt<TElement>(ref TElement[] Source, int Index)
{
var Result = new TElement[Source.Length - 1];
if (Index > 0)
Array.Copy(Source, 0, Result, 0, Index);
if (Index < Source.Length - 1)
Array.Copy(Source, Index + 1, Result, Index, Source.Length - Index - 1);
Source = Result;
}
}
Performance wise, it is decent, but it could probably be improved. Remove relies on IndexOf and a new array is created for each element you wish to remove by calling RemoveAt.
IndexOf is the only extension method as it does not need to return the original array. New accepts multiple elements of some type to produce a new array of said type. All other methods must accept the original array as a reference so there is no need to assign the result afterward as that happens internally already.
I would've defined a Merge method for merging two arrays; however, that can already be accomplished with Add method by passing in an actual array versus multiple, individual elements. Therefore, Add may be used in the following two ways to join two sets of elements:
Arr.Add<string>(ref myArray, "A", "B", "C");
Or
Arr.Add<string>(ref myArray, anotherArray);
I know this article is ten years old and therefore probably dead, but here's what I'd try doing:
Use the IEnumerable.Skip() method, found in System.Linq. It will skip the selected element from the array, and return another copy of the array that only contains everything except the selected object. Then just repeat that for every element you want removed and after that save it to a variable.
For example, if we have an array named "Sample" (of type int[]) with 5 numbers. We want to remove the 2nd one, so trying "Sample.Skip(2);" should return the same array except without the 2nd number.
First step
You need to convert the array into a list, you could write an extension method like this
// Convert An array of string to a list of string
public static List<string> ConnvertArrayToList(this string [] array) {
// DECLARE a list of string and add all element of the array into it
List<string> myList = new List<string>();
foreach( string s in array){
myList.Add(s);
}
return myList;
}
Second step
Write an extension method to convert back the list into an array
// convert a list of string to an array
public static string[] ConvertListToArray(this List<string> list) {
string[] array = new string[list.Capacity];
array = list.Select(i => i.ToString()).ToArray();
return array;
}
Last steps
Write your final method, but remember to remove the element at index before converting back to an array like the code show
public static string[] removeAt(string[] array, int index) {
List<string> myList = array.ConnvertArrayToList();
myList.RemoveAt(index);
return myList.ConvertListToArray();
}
examples codes could be find on my blog, keep tracking.
I would like to generate all permutations of a set (a collection), like so:
Collection: 1, 2, 3
Permutations: {1, 2, 3}
{1, 3, 2}
{2, 1, 3}
{2, 3, 1}
{3, 1, 2}
{3, 2, 1}
This isn't a question of "how", in general, but more about how most efficiently.
Also, I wouldn't want to generate ALL permutations and return them, but only generating a single permutation, at a time, and continuing only if necessary (much like Iterators - which I've tried as well, but turned out to be less efficient).
I've tested many algorithms and approaches and came up with this code, which is most efficient of those I tried:
public static bool NextPermutation<T>(T[] elements) where T : IComparable<T>
{
// More efficient to have a variable instead of accessing a property
var count = elements.Length;
// Indicates whether this is the last lexicographic permutation
var done = true;
// Go through the array from last to first
for (var i = count - 1; i > 0; i--)
{
var curr = elements[i];
// Check if the current element is less than the one before it
if (curr.CompareTo(elements[i - 1]) < 0)
{
continue;
}
// An element bigger than the one before it has been found,
// so this isn't the last lexicographic permutation.
done = false;
// Save the previous (bigger) element in a variable for more efficiency.
var prev = elements[i - 1];
// Have a variable to hold the index of the element to swap
// with the previous element (the to-swap element would be
// the smallest element that comes after the previous element
// and is bigger than the previous element), initializing it
// as the current index of the current item (curr).
var currIndex = i;
// Go through the array from the element after the current one to last
for (var j = i + 1; j < count; j++)
{
// Save into variable for more efficiency
var tmp = elements[j];
// Check if tmp suits the "next swap" conditions:
// Smallest, but bigger than the "prev" element
if (tmp.CompareTo(curr) < 0 && tmp.CompareTo(prev) > 0)
{
curr = tmp;
currIndex = j;
}
}
// Swap the "prev" with the new "curr" (the swap-with element)
elements[currIndex] = prev;
elements[i - 1] = curr;
// Reverse the order of the tail, in order to reset it's lexicographic order
for (var j = count - 1; j > i; j--, i++)
{
var tmp = elements[j];
elements[j] = elements[i];
elements[i] = tmp;
}
// Break since we have got the next permutation
// The reason to have all the logic inside the loop is
// to prevent the need of an extra variable indicating "i" when
// the next needed swap is found (moving "i" outside the loop is a
// bad practice, and isn't very readable, so I preferred not doing
// that as well).
break;
}
// Return whether this has been the last lexicographic permutation.
return done;
}
It's usage would be sending an array of elements, and getting back a boolean indicating whether this was the last lexicographical permutation or not, as well as having the array altered to the next permutation.
Usage example:
var arr = new[] {1, 2, 3};
PrintArray(arr);
while (!NextPermutation(arr))
{
PrintArray(arr);
}
The thing is that I'm not happy with the speed of the code.
Iterating over all permutations of an array of size 11 takes about 4 seconds.
Although it could be considered impressive, since the amount of possible permutations of a set of size 11 is 11! which is nearly 40 million.
Logically, with an array of size 12 it will take about 12 times more time, since 12! is 11! * 12, and with an array of size 13 it will take about 13 times more time than the time it took with size 12, and so on.
So you can easily understand how with an array of size 12 and more, it really takes a very long time to go through all permutations.
And I have a strong hunch that I can somehow cut that time by a lot (without switching to a language other than C# - because compiler optimization really does optimize pretty nicely, and I doubt I could optimize as good, manually, in Assembly).
Does anyone know any other way to get that done faster?
Do you have any idea as to how to make the current algorithm faster?
Note that I don't want to use an external library or service in order to do that - I want to have the code itself and I want it to be as efficient as humanly possible.
This might be what you're looking for.
private static bool NextPermutation(int[] numList)
{
/*
Knuths
1. Find the largest index j such that a[j] < a[j + 1]. If no such index exists, the permutation is the last permutation.
2. Find the largest index l such that a[j] < a[l]. Since j + 1 is such an index, l is well defined and satisfies j < l.
3. Swap a[j] with a[l].
4. Reverse the sequence from a[j + 1] up to and including the final element a[n].
*/
var largestIndex = -1;
for (var i = numList.Length - 2; i >= 0; i--)
{
if (numList[i] < numList[i + 1]) {
largestIndex = i;
break;
}
}
if (largestIndex < 0) return false;
var largestIndex2 = -1;
for (var i = numList.Length - 1 ; i >= 0; i--) {
if (numList[largestIndex] < numList[i]) {
largestIndex2 = i;
break;
}
}
var tmp = numList[largestIndex];
numList[largestIndex] = numList[largestIndex2];
numList[largestIndex2] = tmp;
for (int i = largestIndex + 1, j = numList.Length - 1; i < j; i++, j--) {
tmp = numList[i];
numList[i] = numList[j];
numList[j] = tmp;
}
return true;
}
Update 2018-05-28:
A new multithreaded version (lot faster) is available below as another answer.
Also an article about permutation: Permutations: Fast implementations and a new indexing algorithm allowing multithreading
A little bit too late...
According to recent tests (updated 2018-05-22)
Fastest is mine BUT not in lexicographic order
For fastest lexicographic order, Sani Singh Huttunen solution seems to be the way to go.
Performance test results for 10 items (10!) in release on my machine (millisecs):
Ouellet : 29
SimpleVar: 95
Erez Robinson : 156
Sani Singh Huttunen : 37
Pengyang : 45047
Performance test results for 13 items (13!) in release on my machine (seconds):
Ouellet : 48.437
SimpleVar: 159.869
Erez Robinson : 327.781
Sani Singh Huttunen : 64.839
Advantages of my solution:
Heap's algorithm (Single swap per permutation)
No multiplication (like some implementations seen on the web)
Inlined swap
Generic
No unsafe code
In place (very low memory usage)
No modulo (only first bit compare)
My implementation of Heap's algorithm:
using System;
using System.Collections.Generic;
using System.Diagnostics;
using System.Linq;
using System.Runtime.CompilerServices;
namespace WpfPermutations
{
/// <summary>
/// EO: 2016-04-14
/// Generator of all permutations of an array of anything.
/// Base on Heap's Algorithm. See: https://en.wikipedia.org/wiki/Heap%27s_algorithm#cite_note-3
/// </summary>
public static class Permutations
{
/// <summary>
/// Heap's algorithm to find all pmermutations. Non recursive, more efficient.
/// </summary>
/// <param name="items">Items to permute in each possible ways</param>
/// <param name="funcExecuteAndTellIfShouldStop"></param>
/// <returns>Return true if cancelled</returns>
public static bool ForAllPermutation<T>(T[] items, Func<T[], bool> funcExecuteAndTellIfShouldStop)
{
int countOfItem = items.Length;
if (countOfItem <= 1)
{
return funcExecuteAndTellIfShouldStop(items);
}
var indexes = new int[countOfItem];
// Unecessary. Thanks to NetManage for the advise
// for (int i = 0; i < countOfItem; i++)
// {
// indexes[i] = 0;
// }
if (funcExecuteAndTellIfShouldStop(items))
{
return true;
}
for (int i = 1; i < countOfItem;)
{
if (indexes[i] < i)
{ // On the web there is an implementation with a multiplication which should be less efficient.
if ((i & 1) == 1) // if (i % 2 == 1) ... more efficient ??? At least the same.
{
Swap(ref items[i], ref items[indexes[i]]);
}
else
{
Swap(ref items[i], ref items[0]);
}
if (funcExecuteAndTellIfShouldStop(items))
{
return true;
}
indexes[i]++;
i = 1;
}
else
{
indexes[i++] = 0;
}
}
return false;
}
/// <summary>
/// This function is to show a linq way but is far less efficient
/// From: StackOverflow user: Pengyang : http://stackoverflow.com/questions/756055/listing-all-permutations-of-a-string-integer
/// </summary>
/// <typeparam name="T"></typeparam>
/// <param name="list"></param>
/// <param name="length"></param>
/// <returns></returns>
static IEnumerable<IEnumerable<T>> GetPermutations<T>(IEnumerable<T> list, int length)
{
if (length == 1) return list.Select(t => new T[] { t });
return GetPermutations(list, length - 1)
.SelectMany(t => list.Where(e => !t.Contains(e)),
(t1, t2) => t1.Concat(new T[] { t2 }));
}
/// <summary>
/// Swap 2 elements of same type
/// </summary>
/// <typeparam name="T"></typeparam>
/// <param name="a"></param>
/// <param name="b"></param>
[MethodImpl(MethodImplOptions.AggressiveInlining)]
static void Swap<T>(ref T a, ref T b)
{
T temp = a;
a = b;
b = temp;
}
/// <summary>
/// Func to show how to call. It does a little test for an array of 4 items.
/// </summary>
public static void Test()
{
ForAllPermutation("123".ToCharArray(), (vals) =>
{
Console.WriteLine(String.Join("", vals));
return false;
});
int[] values = new int[] { 0, 1, 2, 4 };
Console.WriteLine("Ouellet heap's algorithm implementation");
ForAllPermutation(values, (vals) =>
{
Console.WriteLine(String.Join("", vals));
return false;
});
Console.WriteLine("Linq algorithm");
foreach (var v in GetPermutations(values, values.Length))
{
Console.WriteLine(String.Join("", v));
}
// Performance Heap's against Linq version : huge differences
int count = 0;
values = new int[10];
for (int n = 0; n < values.Length; n++)
{
values[n] = n;
}
Stopwatch stopWatch = new Stopwatch();
ForAllPermutation(values, (vals) =>
{
foreach (var v in vals)
{
count++;
}
return false;
});
stopWatch.Stop();
Console.WriteLine($"Ouellet heap's algorithm implementation {count} items in {stopWatch.ElapsedMilliseconds} millisecs");
count = 0;
stopWatch.Reset();
stopWatch.Start();
foreach (var vals in GetPermutations(values, values.Length))
{
foreach (var v in vals)
{
count++;
}
}
stopWatch.Stop();
Console.WriteLine($"Linq {count} items in {stopWatch.ElapsedMilliseconds} millisecs");
}
}
}
An this is my test code:
Task.Run(() =>
{
int[] values = new int[12];
for (int n = 0; n < values.Length; n++)
{
values[n] = n;
}
// Eric Ouellet Algorithm
int count = 0;
var stopwatch = new Stopwatch();
stopwatch.Reset();
stopwatch.Start();
Permutations.ForAllPermutation(values, (vals) =>
{
foreach (var v in vals)
{
count++;
}
return false;
});
stopwatch.Stop();
Console.WriteLine($"This {count} items in {stopwatch.ElapsedMilliseconds} millisecs");
// Simple Plan Algorithm
count = 0;
stopwatch.Reset();
stopwatch.Start();
PermutationsSimpleVar permutations2 = new PermutationsSimpleVar();
permutations2.Permutate(1, values.Length, (int[] vals) =>
{
foreach (var v in vals)
{
count++;
}
});
stopwatch.Stop();
Console.WriteLine($"Simple Plan {count} items in {stopwatch.ElapsedMilliseconds} millisecs");
// ErezRobinson Algorithm
count = 0;
stopwatch.Reset();
stopwatch.Start();
foreach(var vals in PermutationsErezRobinson.QuickPerm(values))
{
foreach (var v in vals)
{
count++;
}
};
stopwatch.Stop();
Console.WriteLine($"Erez Robinson {count} items in {stopwatch.ElapsedMilliseconds} millisecs");
});
Usage examples:
ForAllPermutation("123".ToCharArray(), (vals) =>
{
Console.WriteLine(String.Join("", vals));
return false;
});
int[] values = new int[] { 0, 1, 2, 4 };
ForAllPermutation(values, (vals) =>
{
Console.WriteLine(String.Join("", vals));
return false;
});
Well, if you can handle it in C and then translate to your language of choice, you can't really go much faster than this, because the time will be dominated by print:
void perm(char* s, int n, int i){
if (i >= n-1) print(s);
else {
perm(s, n, i+1);
for (int j = i+1; j<n; j++){
swap(s[i], s[j]);
perm(s, n, i+1);
swap(s[i], s[j]);
}
}
}
perm("ABC", 3, 0);
Update 2018-05-28, a new version, the fastest ... (multi-threaded)
Time taken for fastest algorithms
Need: Sani Singh Huttunen (fastest lexico) solution and my new OuelletLexico3 which support indexing
Indexing has 2 main advantages:
allows to get anyone permutation directly
allows multi-threading (derived from the first advantage)
Article: Permutations: Fast implementations and a new indexing algorithm allowing multithreading
On my machine (6 hyperthread cores : 12 threads) Xeon E5-1660 0 # 3.30Ghz, tests algorithms running with empty stuff to do for 13! items (time in millisecs):
53071: Ouellet (implementation of Heap)
65366: Sani Singh Huttunen (Fastest lexico)
11377: Mix OuelletLexico3 - Sani Singh Huttunen
A side note: using shares properties/variables between threads for permutation action will strongly impact performance if their usage is modification (read / write). Doing so will generate "false sharing" between threads. You will not get expected performance. I got this behavior while testing. My experience showed problems when I try to increase the global variable for the total count of permutation.
Usage:
PermutationMixOuelletSaniSinghHuttunen.ExecuteForEachPermutationMT(
new int[] {1, 2, 3, 4},
p =>
{
Console.WriteLine($"Values: {p[0]}, {p[1]}, p[2]}, {p[3]}");
});
Code:
using System;
using System.Runtime.CompilerServices;
namespace WpfPermutations
{
public class Factorial
{
// ************************************************************************
protected static long[] FactorialTable = new long[21];
// ************************************************************************
static Factorial()
{
FactorialTable[0] = 1; // To prevent divide by 0
long f = 1;
for (int i = 1; i <= 20; i++)
{
f = f * i;
FactorialTable[i] = f;
}
}
// ************************************************************************
[MethodImpl(MethodImplOptions.AggressiveInlining)]
public static long GetFactorial(int val) // a long can only support up to 20!
{
if (val > 20)
{
throw new OverflowException($"{nameof(Factorial)} only support a factorial value <= 20");
}
return FactorialTable[val];
}
// ************************************************************************
}
}
namespace WpfPermutations
{
public class PermutationSaniSinghHuttunen
{
public static bool NextPermutation(int[] numList)
{
/*
Knuths
1. Find the largest index j such that a[j] < a[j + 1]. If no such index exists, the permutation is the last permutation.
2. Find the largest index l such that a[j] < a[l]. Since j + 1 is such an index, l is well defined and satisfies j < l.
3. Swap a[j] with a[l].
4. Reverse the sequence from a[j + 1] up to and including the final element a[n].
*/
var largestIndex = -1;
for (var i = numList.Length - 2; i >= 0; i--)
{
if (numList[i] < numList[i + 1])
{
largestIndex = i;
break;
}
}
if (largestIndex < 0) return false;
var largestIndex2 = -1;
for (var i = numList.Length - 1; i >= 0; i--)
{
if (numList[largestIndex] < numList[i])
{
largestIndex2 = i;
break;
}
}
var tmp = numList[largestIndex];
numList[largestIndex] = numList[largestIndex2];
numList[largestIndex2] = tmp;
for (int i = largestIndex + 1, j = numList.Length - 1; i < j; i++, j--)
{
tmp = numList[i];
numList[i] = numList[j];
numList[j] = tmp;
}
return true;
}
}
}
using System;
namespace WpfPermutations
{
public class PermutationOuelletLexico3<T> // Enable indexing
{
// ************************************************************************
private T[] _sortedValues;
private bool[] _valueUsed;
public readonly long MaxIndex; // long to support 20! or less
// ************************************************************************
public PermutationOuelletLexico3(T[] sortedValues)
{
_sortedValues = sortedValues;
Result = new T[_sortedValues.Length];
_valueUsed = new bool[_sortedValues.Length];
MaxIndex = Factorial.GetFactorial(_sortedValues.Length);
}
// ************************************************************************
public T[] Result { get; private set; }
// ************************************************************************
/// <summary>
/// Sort Index is 0 based and should be less than MaxIndex. Otherwise you get an exception.
/// </summary>
/// <param name="sortIndex"></param>
/// <param name="result">Value is not used as inpu, only as output. Re-use buffer in order to save memory</param>
/// <returns></returns>
public void GetSortedValuesFor(long sortIndex)
{
int size = _sortedValues.Length;
if (sortIndex < 0)
{
throw new ArgumentException("sortIndex should greater or equal to 0.");
}
if (sortIndex >= MaxIndex)
{
throw new ArgumentException("sortIndex should less than factorial(the lenght of items)");
}
for (int n = 0; n < _valueUsed.Length; n++)
{
_valueUsed[n] = false;
}
long factorielLower = MaxIndex;
for (int index = 0; index < size; index++)
{
long factorielBigger = factorielLower;
factorielLower = Factorial.GetFactorial(size - index - 1); // factorielBigger / inverseIndex;
int resultItemIndex = (int)(sortIndex % factorielBigger / factorielLower);
int correctedResultItemIndex = 0;
for(;;)
{
if (! _valueUsed[correctedResultItemIndex])
{
resultItemIndex--;
if (resultItemIndex < 0)
{
break;
}
}
correctedResultItemIndex++;
}
Result[index] = _sortedValues[correctedResultItemIndex];
_valueUsed[correctedResultItemIndex] = true;
}
}
// ************************************************************************
}
}
using System;
using System.Collections.Generic;
using System.Threading.Tasks;
namespace WpfPermutations
{
public class PermutationMixOuelletSaniSinghHuttunen
{
// ************************************************************************
private long _indexFirst;
private long _indexLastExclusive;
private int[] _sortedValues;
// ************************************************************************
public PermutationMixOuelletSaniSinghHuttunen(int[] sortedValues, long indexFirst = -1, long indexLastExclusive = -1)
{
if (indexFirst == -1)
{
indexFirst = 0;
}
if (indexLastExclusive == -1)
{
indexLastExclusive = Factorial.GetFactorial(sortedValues.Length);
}
if (indexFirst >= indexLastExclusive)
{
throw new ArgumentException($"{nameof(indexFirst)} should be less than {nameof(indexLastExclusive)}");
}
_indexFirst = indexFirst;
_indexLastExclusive = indexLastExclusive;
_sortedValues = sortedValues;
}
// ************************************************************************
public void ExecuteForEachPermutation(Action<int[]> action)
{
// Console.WriteLine($"Thread {System.Threading.Thread.CurrentThread.ManagedThreadId} started: {_indexFirst} {_indexLastExclusive}");
long index = _indexFirst;
PermutationOuelletLexico3<int> permutationOuellet = new PermutationOuelletLexico3<int>(_sortedValues);
permutationOuellet.GetSortedValuesFor(index);
action(permutationOuellet.Result);
index++;
int[] values = permutationOuellet.Result;
while (index < _indexLastExclusive)
{
PermutationSaniSinghHuttunen.NextPermutation(values);
action(values);
index++;
}
// Console.WriteLine($"Thread {System.Threading.Thread.CurrentThread.ManagedThreadId} ended: {DateTime.Now.ToString("yyyyMMdd_HHmmss_ffffff")}");
}
// ************************************************************************
public static void ExecuteForEachPermutationMT(int[] sortedValues, Action<int[]> action)
{
int coreCount = Environment.ProcessorCount; // Hyper treading are taken into account (ex: on a 4 cores hyperthreaded = 8)
long itemsFactorial = Factorial.GetFactorial(sortedValues.Length);
long partCount = (long)Math.Ceiling((double)itemsFactorial / (double)coreCount);
long startIndex = 0;
var tasks = new List<Task>();
for (int coreIndex = 0; coreIndex < coreCount; coreIndex++)
{
long stopIndex = Math.Min(startIndex + partCount, itemsFactorial);
PermutationMixOuelletSaniSinghHuttunen mix = new PermutationMixOuelletSaniSinghHuttunen(sortedValues, startIndex, stopIndex);
Task task = Task.Run(() => mix.ExecuteForEachPermutation(action));
tasks.Add(task);
if (stopIndex == itemsFactorial)
{
break;
}
startIndex = startIndex + partCount;
}
Task.WaitAll(tasks.ToArray());
}
// ************************************************************************
}
}
The fastest permutation algorithm that i know of is the QuickPerm algorithm.
Here is the implementation, it uses yield return so you can iterate one at a time like required.
Code:
public static IEnumerable<IEnumerable<T>> QuickPerm<T>(this IEnumerable<T> set)
{
int N = set.Count();
int[] a = new int[N];
int[] p = new int[N];
var yieldRet = new T[N];
List<T> list = new List<T>(set);
int i, j, tmp; // Upper Index i; Lower Index j
for (i = 0; i < N; i++)
{
// initialize arrays; a[N] can be any type
a[i] = i + 1; // a[i] value is not revealed and can be arbitrary
p[i] = 0; // p[i] == i controls iteration and index boundaries for i
}
yield return list;
//display(a, 0, 0); // remove comment to display array a[]
i = 1; // setup first swap points to be 1 and 0 respectively (i & j)
while (i < N)
{
if (p[i] < i)
{
j = i%2*p[i]; // IF i is odd then j = p[i] otherwise j = 0
tmp = a[j]; // swap(a[j], a[i])
a[j] = a[i];
a[i] = tmp;
//MAIN!
for (int x = 0; x < N; x++)
{
yieldRet[x] = list[a[x]-1];
}
yield return yieldRet;
//display(a, j, i); // remove comment to display target array a[]
// MAIN!
p[i]++; // increase index "weight" for i by one
i = 1; // reset index i to 1 (assumed)
}
else
{
// otherwise p[i] == i
p[i] = 0; // reset p[i] to zero
i++; // set new index value for i (increase by one)
} // if (p[i] < i)
} // while(i < N)
}
Here is the fastest implementation I ended up with:
public class Permutations
{
private readonly Mutex _mutex = new Mutex();
private Action<int[]> _action;
private Action<IntPtr> _actionUnsafe;
private unsafe int* _arr;
private IntPtr _arrIntPtr;
private unsafe int* _last;
private unsafe int* _lastPrev;
private unsafe int* _lastPrevPrev;
public int Size { get; private set; }
public bool IsRunning()
{
return this._mutex.SafeWaitHandle.IsClosed;
}
public bool Permutate(int start, int count, Action<int[]> action, bool async = false)
{
return this.Permutate(start, count, action, null, async);
}
public bool Permutate(int start, int count, Action<IntPtr> actionUnsafe, bool async = false)
{
return this.Permutate(start, count, null, actionUnsafe, async);
}
private unsafe bool Permutate(int start, int count, Action<int[]> action, Action<IntPtr> actionUnsafe, bool async = false)
{
if (!this._mutex.WaitOne(0))
{
return false;
}
var x = (Action)(() =>
{
this._actionUnsafe = actionUnsafe;
this._action = action;
this.Size = count;
this._arr = (int*)Marshal.AllocHGlobal(count * sizeof(int));
this._arrIntPtr = new IntPtr(this._arr);
for (var i = 0; i < count - 3; i++)
{
this._arr[i] = start + i;
}
this._last = this._arr + count - 1;
this._lastPrev = this._last - 1;
this._lastPrevPrev = this._lastPrev - 1;
*this._last = count - 1;
*this._lastPrev = count - 2;
*this._lastPrevPrev = count - 3;
this.Permutate(count, this._arr);
});
if (!async)
{
x();
}
else
{
new Thread(() => x()).Start();
}
return true;
}
private unsafe void Permutate(int size, int* start)
{
if (size == 3)
{
this.DoAction();
Swap(this._last, this._lastPrev);
this.DoAction();
Swap(this._last, this._lastPrevPrev);
this.DoAction();
Swap(this._last, this._lastPrev);
this.DoAction();
Swap(this._last, this._lastPrevPrev);
this.DoAction();
Swap(this._last, this._lastPrev);
this.DoAction();
return;
}
var sizeDec = size - 1;
var startNext = start + 1;
var usedStarters = 0;
for (var i = 0; i < sizeDec; i++)
{
this.Permutate(sizeDec, startNext);
usedStarters |= 1 << *start;
for (var j = startNext; j <= this._last; j++)
{
var mask = 1 << *j;
if ((usedStarters & mask) != mask)
{
Swap(start, j);
break;
}
}
}
this.Permutate(sizeDec, startNext);
if (size == this.Size)
{
this._mutex.ReleaseMutex();
}
}
private unsafe void DoAction()
{
if (this._action == null)
{
if (this._actionUnsafe != null)
{
this._actionUnsafe(this._arrIntPtr);
}
return;
}
var result = new int[this.Size];
fixed (int* pt = result)
{
var limit = pt + this.Size;
var resultPtr = pt;
var arrayPtr = this._arr;
while (resultPtr < limit)
{
*resultPtr = *arrayPtr;
resultPtr++;
arrayPtr++;
}
}
this._action(result);
}
private static unsafe void Swap(int* a, int* b)
{
var tmp = *a;
*a = *b;
*b = tmp;
}
}
Usage and testing performance:
var perms = new Permutations();
var sw1 = Stopwatch.StartNew();
perms.Permutate(0,
11,
(Action<int[]>)null); // Comment this line and...
//PrintArr); // Uncomment this line, to print permutations
sw1.Stop();
Console.WriteLine(sw1.Elapsed);
Printing method:
private static void PrintArr(int[] arr)
{
Console.WriteLine(string.Join(",", arr));
}
Going deeper:
I did not even think about this for a very long time, so I can only explain my code so much, but here's the general idea:
Permutations aren't lexicographic - this allows me to practically perform less operations between permutations.
The implementation is recursive, and when the "view" size is 3, it skips the complex logic and just performs 6 swaps to get the 6 permutations (or sub-permutations, if you will).
Because the permutations aren't in a lexicographic order, how can I decide which element to bring to the start of the current "view" (sub permutation)? I keep record of elements that were already used as "starters" in the current sub-permutation recursive call and simply search linearly for one that wasn't used in the tail of my array.
The implementation is for integers only, so to permute over a generic collection of elements you simply use the Permutations class to permute indices instead of your actual collection.
The Mutex is there just to ensure things don't get screwed when the execution is asynchronous (notice that you can pass an UnsafeAction parameter that will in turn get a pointer to the permuted array. You must not change the order of elements in that array (pointer)! If you want to, you should copy the array to a tmp array or just use the safe action parameter which takes care of that for you - the passed array is already a copy).
Note:
I have no idea how good this implementation really is - I haven't touched it in so long.
Test and compare to other implementations on your own, and let me know if you have any feedback!
Enjoy.
Here is a generic permutation finder that will iterate through every permutation of a collection and call an evalution function. If the evalution function returns true (it found the answer it was looking for), the permutation finder stops processing.
public class PermutationFinder<T>
{
private T[] items;
private Predicate<T[]> SuccessFunc;
private bool success = false;
private int itemsCount;
public void Evaluate(T[] items, Predicate<T[]> SuccessFunc)
{
this.items = items;
this.SuccessFunc = SuccessFunc;
this.itemsCount = items.Count();
Recurse(0);
}
private void Recurse(int index)
{
T tmp;
if (index == itemsCount)
success = SuccessFunc(items);
else
{
for (int i = index; i < itemsCount; i++)
{
tmp = items[index];
items[index] = items[i];
items[i] = tmp;
Recurse(index + 1);
if (success)
break;
tmp = items[index];
items[index] = items[i];
items[i] = tmp;
}
}
}
}
Here is a simple implementation:
class Program
{
static void Main(string[] args)
{
new Program().Start();
}
void Start()
{
string[] items = new string[5];
items[0] = "A";
items[1] = "B";
items[2] = "C";
items[3] = "D";
items[4] = "E";
new PermutationFinder<string>().Evaluate(items, Evaluate);
Console.ReadLine();
}
public bool Evaluate(string[] items)
{
Console.WriteLine(string.Format("{0},{1},{2},{3},{4}", items[0], items[1], items[2], items[3], items[4]));
bool someCondition = false;
if (someCondition)
return true; // Tell the permutation finder to stop.
return false;
}
}
Here is a recursive implementation with complexity O(n * n!)1 based on swapping of the elements of an array. The array is initialised with values from 1, 2, ..., n.
using System;
namespace Exercise
{
class Permutations
{
static void Main(string[] args)
{
int setSize = 3;
FindPermutations(setSize);
}
//-----------------------------------------------------------------------------
/* Method: FindPermutations(n) */
private static void FindPermutations(int n)
{
int[] arr = new int[n];
for (int i = 0; i < n; i++)
{
arr[i] = i + 1;
}
int iEnd = arr.Length - 1;
Permute(arr, iEnd);
}
//-----------------------------------------------------------------------------
/* Method: Permute(arr) */
private static void Permute(int[] arr, int iEnd)
{
if (iEnd == 0)
{
PrintArray(arr);
return;
}
Permute(arr, iEnd - 1);
for (int i = 0; i < iEnd; i++)
{
swap(ref arr[i], ref arr[iEnd]);
Permute(arr, iEnd - 1);
swap(ref arr[i], ref arr[iEnd]);
}
}
}
}
On each recursive step we swap the last element with the current element pointed to by the local variable in the for loop and then we indicate the uniqueness of the swapping by: incrementing the local variable of the for loop and decrementing the termination condition of the for loop, which is initially set to the number of the elements in the array, when the latter becomes zero we terminate the recursion.
Here are the helper functions:
//-----------------------------------------------------------------------------
/*
Method: PrintArray()
*/
private static void PrintArray(int[] arr, string label = "")
{
Console.WriteLine(label);
Console.Write("{");
for (int i = 0; i < arr.Length; i++)
{
Console.Write(arr[i]);
if (i < arr.Length - 1)
{
Console.Write(", ");
}
}
Console.WriteLine("}");
}
//-----------------------------------------------------------------------------
/*
Method: swap(ref int a, ref int b)
*/
private static void swap(ref int a, ref int b)
{
int temp = a;
a = b;
b = temp;
}
1. There are n! permutations of n elements to be printed.
I would be surprised if there are really order of magnitude improvements to be found. If there are, then C# needs fundamental improvement. Furthermore doing anything interesting with your permutation will generally take more work than generating it. So the cost of generating is going to be insignificant in the overall scheme of things.
That said, I would suggest trying the following things. You have already tried iterators. But have you tried having a function that takes a closure as input, then then calls that closure for each permutation found? Depending on internal mechanics of C#, this may be faster.
Similarly, have you tried having a function that returns a closure that will iterate over a specific permutation?
With either approach, there are a number of micro-optimizations you can experiment with. For instance you can sort your input array, and after that you always know what order it is in. For example you can have an array of bools indicating whether that element is less than the next one, and rather than do comparisons, you can just look at that array.
There's an accessible introduction to the algorithms and survey of implementations in Steven Skiena's Algorithm Design Manual (chapter 14.4 in the second edition)
Skiena references D. Knuth. The Art of Computer Programming, Volume 4 Fascicle 2: Generating All Tuples and Permutations. Addison Wesley, 2005.
I created an algorithm slightly faster than Knuth's one:
11 elements:
mine: 0.39 seconds
Knuth's: 0.624 seconds
13 elements:
mine: 56.615 seconds
Knuth's: 98.681 seconds
Here's my code in Java:
public static void main(String[] args)
{
int n=11;
int a,b,c,i,tmp;
int end=(int)Math.floor(n/2);
int[][] pos = new int[end+1][2];
int[] perm = new int[n];
for(i=0;i<n;i++) perm[i]=i;
while(true)
{
//this is where you can use the permutations (perm)
i=0;
c=n;
while(pos[i][1]==c-2 && pos[i][0]==c-1)
{
pos[i][0]=0;
pos[i][1]=0;
i++;
c-=2;
}
if(i==end) System.exit(0);
a=(pos[i][0]+1)%c+i;
b=pos[i][0]+i;
tmp=perm[b];
perm[b]=perm[a];
perm[a]=tmp;
if(pos[i][0]==c-1)
{
pos[i][0]=0;
pos[i][1]++;
}
else
{
pos[i][0]++;
}
}
}
The problem is my algorithm only works for odd numbers of elements. I wrote this code quickly so I'm pretty sure there's a better way to implement my idea to get better performance, but I don't really have the time to work on it right now to optimize it and solve the issue when the number of elements is even.
It's one swap for every permutation and it uses a really simple way to know which elements to swap.
I wrote an explanation of the method behind the code on my blog: http://antoinecomeau.blogspot.ca/2015/01/fast-generation-of-all-permutations.html
As the author of this question was asking about an algorithm:
[...] generating a single permutation, at a time, and continuing only if necessary
I would suggest considering Steinhaus–Johnson–Trotter algorithm.
Steinhaus–Johnson–Trotter algorithm on Wikipedia
Beautifully explained here
It's 1 am and I was watching TV and thought of this same question, but with string values.
Given a word find all permutations. You can easily modify this to handle an array, sets, etc.
Took me a bit to work it out, but the solution I came up was this:
string word = "abcd";
List<string> combinations = new List<string>();
for(int i=0; i<word.Length; i++)
{
for (int j = 0; j < word.Length; j++)
{
if (i < j)
combinations.Add(word[i] + word.Substring(j) + word.Substring(0, i) + word.Substring(i + 1, j - (i + 1)));
else if (i > j)
{
if(i== word.Length -1)
combinations.Add(word[i] + word.Substring(0, i));
else
combinations.Add(word[i] + word.Substring(0, i) + word.Substring(i + 1));
}
}
}
Here's the same code as above, but with some comments
string word = "abcd";
List<string> combinations = new List<string>();
//i is the first letter of the new word combination
for(int i=0; i<word.Length; i++)
{
for (int j = 0; j < word.Length; j++)
{
//add the first letter of the word, j is past i so we can get all the letters from j to the end
//then add all the letters from the front to i, then skip over i (since we already added that as the beginning of the word)
//and get the remaining letters from i+1 to right before j.
if (i < j)
combinations.Add(word[i] + word.Substring(j) + word.Substring(0, i) + word.Substring(i + 1, j - (i + 1)));
else if (i > j)
{
//if we're at the very last word no need to get the letters after i
if(i== word.Length -1)
combinations.Add(word[i] + word.Substring(0, i));
//add i as the first letter of the word, then get all the letters up to i, skip i, and then add all the lettes after i
else
combinations.Add(word[i] + word.Substring(0, i) + word.Substring(i + 1));
}
}
}
//+------------------------------------------------------------------+
//| |
//+------------------------------------------------------------------+
/**
* http://marknelson.us/2002/03/01/next-permutation/
* Rearranges the elements into the lexicographically next greater permutation and returns true.
* When there are no more greater permutations left, the function eventually returns false.
*/
// next lexicographical permutation
template <typename T>
bool next_permutation(T &arr[], int firstIndex, int lastIndex)
{
int i = lastIndex;
while (i > firstIndex)
{
int ii = i--;
T curr = arr[i];
if (curr < arr[ii])
{
int j = lastIndex;
while (arr[j] <= curr) j--;
Swap(arr[i], arr[j]);
while (ii < lastIndex)
Swap(arr[ii++], arr[lastIndex--]);
return true;
}
}
return false;
}
//+------------------------------------------------------------------+
//| |
//+------------------------------------------------------------------+
/**
* Swaps two variables or two array elements.
* using references/pointers to speed up swapping.
*/
template<typename T>
void Swap(T &var1, T &var2)
{
T temp;
temp = var1;
var1 = var2;
var2 = temp;
}
//+------------------------------------------------------------------+
//| |
//+------------------------------------------------------------------+
// driver program to test above function
#define N 3
void OnStart()
{
int i, x[N];
for (i = 0; i < N; i++) x[i] = i + 1;
printf("The %i! possible permutations with %i elements:", N, N);
do
{
printf("%s", ArrayToString(x));
} while (next_permutation(x, 0, N - 1));
}
// Output:
// The 3! possible permutations with 3 elements:
// "1,2,3"
// "1,3,2"
// "2,1,3"
// "2,3,1"
// "3,1,2"
// "3,2,1"
// Permutations are the different ordered arrangements of an n-element
// array. An n-element array has exactly n! full-length permutations.
// This iterator object allows to iterate all full length permutations
// one by one of an array of n distinct elements.
// The iterator changes the given array in-place.
// Permutations('ABCD') => ABCD DBAC ACDB DCBA
// BACD BDAC CADB CDBA
// CABD ADBC DACB BDCA
// ACBD DABC ADCB DBCA
// BCAD BADC CDAB CBDA
// CBAD ABDC DCAB BCDA
// count of permutations = n!
// Heap's algorithm (Single swap per permutation)
// http://www.quickperm.org/quickperm.php
// https://stackoverflow.com/a/36634935/4208440
// https://en.wikipedia.org/wiki/Heap%27s_algorithm
// My implementation of Heap's algorithm:
template<typename T>
class PermutationsIterator
{
int b, e, n;
int c[32]; /* control array: mixed radix number in rising factorial base.
the i-th digit has base i, which means that the digit must be
strictly less than i. The first digit is always 0, the second
can be 0 or 1, the third 0, 1 or 2, and so on.
ArrayResize isn't strictly necessary, int c[32] would suffice
for most practical purposes. Also, it is much faster */
public:
PermutationsIterator(T &arr[], int firstIndex, int lastIndex)
{
this.b = firstIndex; // v.begin()
this.e = lastIndex; // v.end()
this.n = e - b + 1;
ArrayInitialize(c, 0);
}
// Rearranges the input array into the next permutation and returns true.
// When there are no more permutations left, the function returns false.
bool next(T &arr[])
{
// find index to update
int i = 1;
// reset all the previous indices that reached the maximum possible values
while (c[i] == i)
{
c[i] = 0;
++i;
}
// no more permutations left
if (i == n)
return false;
// generate next permutation
int j = (i & 1) == 1 ? c[i] : 0; // IF i is odd then j = c[i] otherwise j = 0.
swap(arr[b + j], arr[b + i]); // generate a new permutation from previous permutation using a single swap
// Increment that index
++c[i];
return true;
}
};
I found this algo on rosetta code and it is really the fastest one I tried. http://rosettacode.org/wiki/Permutations#C
/* Boothroyd method; exactly N! swaps, about as fast as it gets */
void boothroyd(int *x, int n, int nn, int callback(int *, int))
{
int c = 0, i, t;
while (1) {
if (n > 2) boothroyd(x, n - 1, nn, callback);
if (c >= n - 1) return;
i = (n & 1) ? 0 : c;
c++;
t = x[n - 1], x[n - 1] = x[i], x[i] = t;
if (callback) callback(x, nn);
}
}
/* entry for Boothroyd method */
void perm2(int *x, int n, int callback(int*, int))
{
if (callback) callback(x, n);
boothroyd(x, n, n, callback);
}
If you just want to calculate the number of possible permutations you can avoid all that hard work above and use something like this (contrived in c#):
public static class ContrivedUtils
{
public static Int64 Permutations(char[] array)
{
if (null == array || array.Length == 0) return 0;
Int64 permutations = array.Length;
for (var pos = permutations; pos > 1; pos--)
permutations *= pos - 1;
return permutations;
}
}
You call it like this:
var permutations = ContrivedUtils.Permutations("1234".ToCharArray());
// output is: 24
var permutations = ContrivedUtils.Permutations("123456789".ToCharArray());
// output is: 362880
Simple C# recursive solution by swapping, for the initial call the index must be 0
static public void Permute<T>(List<T> input, List<List<T>> permutations, int index)
{
if (index == input.Count - 1)
{
permutations.Add(new List<T>(input));
return;
}
Permute(input, permutations, index + 1);
for (int i = index+1 ; i < input.Count; i++)
{
//swap
T temp = input[index];
input[index] = input[i];
input[i] = temp;
Permute(input, permutations, index + 1);
//swap back
temp = input[index];
input[index] = input[i];
input[i] = temp;
}
}
How can I change my C# code below to list all possible permutations without repetitions? For example: The result of 2 dice rolls would produce 1,1,2 so that means 2,1,1 should not appear.
Below is my code:
string[] Permutate(int input)
{
string[] dice;
int numberOfDice = input;
const int diceFace = 6;
dice = new string[(int)Math.Pow(diceFace, numberOfDice)];
int indexNumber = (int)Math.Pow(diceFace, numberOfDice);
int range = (int)Math.Pow(diceFace, numberOfDice) / 6;
int diceNumber = 1;
int counter = 0;
for (int i = 1; i <= indexNumber; i++)
{
if (range != 0)
{
dice[i - 1] += diceNumber + " ";
counter++;
if (counter == range)
{
counter = 0;
diceNumber++;
}
if (i == indexNumber)
{
range /= 6;
i = 0;
}
if (diceNumber == 7)
{
diceNumber = 1;
}
}
Thread.Sleep(1);
}
return dice;
}
The simplest possible way I could think of:
List<string> dices = new List<string>();
for (int i = 1; i <= 6; i++)
{
for (int j = i; j <= 6; j++)
{
for (int k = j; k <= 6; k++)
{
dices.Add(string.Format("{0} {1} {2}", i, j, k));
}
}
}
I have written a class to handle common functions for working with the binomial coefficient, which is the type of problem that your problem falls under. It performs the following tasks:
Outputs all the K-indexes in a nice format for any N choose K to a file. The K-indexes can be substituted with more descriptive strings or letters. This method makes solving this type of problem quite trivial.
Converts the K-indexes to the proper index of an entry in the sorted binomial coefficient table. This technique is much faster than older published techniques that rely on iteration. It does this by using a mathematical property inherent in Pascal's Triangle. My paper talks about this. I believe I am the first to discover and publish this technique, but I could be wrong.
Converts the index in a sorted binomial coefficient table to the corresponding K-indexes.
Uses Mark Dominus method to calculate the binomial coefficient, which is much less likely to overflow and works with larger numbers.
The class is written in .NET C# and provides a way to manage the objects related to the problem (if any) by using a generic list. The constructor of this class takes a bool value called InitTable that when true will create a generic list to hold the objects to be managed. If this value is false, then it will not create the table. The table does not need to be created in order to perform the 4 above methods. Accessor methods are provided to access the table.
There is an associated test class which shows how to use the class and its methods. It has been extensively tested with 2 cases and there are no known bugs.
To read about this class and download the code, see Tablizing The Binomial Coeffieicent.
I'm bad at math as well, this may or may not be helpful...
Program.cs
namespace Permutation
{
using System;
using System.Collections.Generic;
class Program
{
static void Main(string[] args)
{
Console.WriteLine("Generating list.");
var dice = new List<ThreeDice>();
for (int x = 1; x <= 6; x++)
{
for (int y = 1; y <= 6; y++)
{
for (int z = 1; z <= 6; z++)
{
var die = new ThreeDice(x, y, z);
if (dice.Contains(die))
{
Console.WriteLine(die + " already exists.");
}
else
{
dice.Add(die);
}
}
}
}
Console.WriteLine(dice.Count + " permutations generated.");
foreach (var die in dice)
{
Console.WriteLine(die);
}
Console.ReadKey();
}
}
}
ThreeDice.cs
namespace Permutation
{
using System;
using System.Collections.Generic;
public class ThreeDice : IEquatable<ThreeDice>
{
public ThreeDice(int dice1, int dice2, int dice3)
{
this.Dice = new int[3];
this.Dice[0] = dice1;
this.Dice[1] = dice2;
this.Dice[2] = dice3;
}
public int[] Dice { get; private set; }
// IEquatable implements this method. List.Contains() will use this method to see if there's a match.
public bool Equals(ThreeDice other)
{
// Get the current dice values into a list.
var currentDice = new List<int>(this.Dice);
// Check to see if the same values exist by removing them one by one.
foreach (int die in other.Dice)
{
currentDice.Remove(die);
}
// If the list is empty, we have a match.
return currentDice.Count == 0;
}
public override string ToString()
{
return "<" + this.Dice[0] + "," + this.Dice[1] + "," + this.Dice[2] + ">";
}
}
}
Good luck.
The important part of the question is that you want distinct sets (regardless of order). So for example, a dice roll of [1, 2, 1] is equal to a dice roll of [1, 1, 2].
I'm sure there are a number of ways to skin this cat, but the first thought that comes to mind is to create a EqualityComparer which will compare the list of dice in the way you want, and then use LINQ with the Distinct() method.
Here is the EqualityComparer, which takes 2 List<int> and says they are equal if the elements are all equal (regardless of order):
private class ListComparer : EqualityComparer<List<int>>
{
public override bool Equals(List<int> x, List<int> y)
{
if (x.Count != y.Count)
return false;
x.Sort();
y.Sort();
for (int i = 0; i < x.Count; i++)
{
if (x[i] != y[i])
return false;
}
return true;
}
public override int GetHashCode(List<int> list)
{
int hc = 0;
foreach (var i in list)
hc ^= i;
return hc;
}
}
And here is the code that uses it. I'm using LINQ to build up the list of all combinations... you could also do this with nested for loops but I like this better for some reason:
public static void Main()
{
var values = new[] { 1,2,3,4,5,6 };
var allCombos = from x in values
from y in values
from z in values
select new List<int>{ x, y, z };
var distinctCombos = allCombos.Distinct(new ListComparer());
Console.WriteLine("#All combos: {0}", allCombos.Count());
Console.WriteLine("#Distinct combos: {0}", distinctCombos.Count());
foreach (var combo in distinctCombos)
Console.WriteLine("{0},{1},{2}", combo[0], combo[1], combo[2]);
}
Hope that helps!
Here is generic c# version using recursion (basically the recursive method takes number of dices or number of times the dice has been tossed) and returns all the combinations strings ( for ex, for '3' as per the question - there will be 56 such combinations).
public string[] GetDiceCombinations(int noOfDicesOrnoOfTossesOfDice)
{
noOfDicesOrnoOfTossesOfDice.Throw("noOfDicesOrnoOfTossesOfDice",
n => n <= 0);
List<string> values = new List<string>();
this.GetDiceCombinations_Recursive(noOfDicesOrnoOfTossesOfDice, 1, "",
values);
return values.ToArray();
}
private void GetDiceCombinations_Recursive(int size, int index, string currentValue,
List<string> values)
{
if (currentValue.Length == size)
{
values.Add(currentValue);
return;
}
for (int i = index; i <= 6; i++)
{
this.GetDiceCombinations_Recursive(size, i, currentValue + i, values);
}
}
Below are corresponding tests...
[TestMethod]
public void Dice_Tests()
{
int[] cOut = new int[] { 6, 21, 56, 126 };
for(int i = 1; i<=4; i++)
{
var c = this.GetDiceCombinations(i);
Assert.AreEqual(cOut[i - 1], c.Length);
}
}
I have an array of Foo objects. How do I remove the second element of the array?
I need something similar to RemoveAt() but for a regular array.
If you don't want to use List:
var foos = new List<Foo>(array);
foos.RemoveAt(index);
return foos.ToArray();
You could try this extension method that I haven't actually tested:
public static T[] RemoveAt<T>(this T[] source, int index)
{
T[] dest = new T[source.Length - 1];
if( index > 0 )
Array.Copy(source, 0, dest, 0, index);
if( index < source.Length - 1 )
Array.Copy(source, index + 1, dest, index, source.Length - index - 1);
return dest;
}
And use it like:
Foo[] bar = GetFoos();
bar = bar.RemoveAt(2);
The nature of arrays is that their length is immutable. You can't add or delete any of the array items.
You will have to create a new array that is one element shorter and copy the old items to the new array, excluding the element you want to delete.
So it is probably better to use a List instead of an array.
I use this method for removing an element from an object array. In my situation, my arrays are small in length. So if you have large arrays you may need another solution.
private int[] RemoveIndices(int[] IndicesArray, int RemoveAt)
{
int[] newIndicesArray = new int[IndicesArray.Length - 1];
int i = 0;
int j = 0;
while (i < IndicesArray.Length)
{
if (i != RemoveAt)
{
newIndicesArray[j] = IndicesArray[i];
j++;
}
i++;
}
return newIndicesArray;
}
LINQ one-line solution:
myArray = myArray.Where((source, index) => index != 1).ToArray();
The 1 in that example is the index of the element to remove -- in this example, per the original question, the 2nd element (with 1 being the second element in C# zero-based array indexing).
A more complete example:
string[] myArray = { "a", "b", "c", "d", "e" };
int indexToRemove = 1;
myArray = myArray.Where((source, index) => index != indexToRemove).ToArray();
After running that snippet, the value of myArray will be { "a", "c", "d", "e" }.
This is a way to delete an array element, as of .Net 3.5, without copying to another array - using the same array instance with Array.Resize<T>:
public static void RemoveAt<T>(ref T[] arr, int index)
{
for (int a = index; a < arr.Length - 1; a++)
{
// moving elements downwards, to fill the gap at [index]
arr[a] = arr[a + 1];
}
// finally, let's decrement Array's size by one
Array.Resize(ref arr, arr.Length - 1);
}
Here is an old version I have that works on version 1.0 of the .NET framework and does not need generic types.
public static Array RemoveAt(Array source, int index)
{
if (source == null)
throw new ArgumentNullException("source");
if (0 > index || index >= source.Length)
throw new ArgumentOutOfRangeException("index", index, "index is outside the bounds of source array");
Array dest = Array.CreateInstance(source.GetType().GetElementType(), source.Length - 1);
Array.Copy(source, 0, dest, 0, index);
Array.Copy(source, index + 1, dest, index, source.Length - index - 1);
return dest;
}
This is used like this:
class Program
{
static void Main(string[] args)
{
string[] x = new string[20];
for (int i = 0; i < x.Length; i++)
x[i] = (i+1).ToString();
string[] y = (string[])MyArrayFunctions.RemoveAt(x, 3);
for (int i = 0; i < y.Length; i++)
Console.WriteLine(y[i]);
}
}
Not exactly the way to go about this, but if the situation is trivial and you value your time, you can try this for nullable types.
Foos[index] = null
and later check for null entries in your logic..
Try below code:
myArray = myArray.Where(s => (myArray.IndexOf(s) != indexValue)).ToArray();
or
myArray = myArray.Where(s => (s != "not_this")).ToArray();
As usual, I'm late to the party...
I'd like to add another option to the nice solutions list already present. =)
I would see this as a good opportunity for Extensions.
Reference:
http://msdn.microsoft.com/en-us/library/bb311042.aspx
So, we define some static class and in it, our Method.
After that, we can use our extended method willy-nilly. =)
using System;
namespace FunctionTesting {
// The class doesn't matter, as long as it's static
public static class SomeRandomClassWhoseNameDoesntMatter {
// Here's the actual method that extends arrays
public static T[] RemoveAt<T>( this T[] oArray, int idx ) {
T[] nArray = new T[oArray.Length - 1];
for( int i = 0; i < nArray.Length; ++i ) {
nArray[i] = ( i < idx ) ? oArray[i] : oArray[i + 1];
}
return nArray;
}
}
// Sample usage...
class Program {
static void Main( string[] args ) {
string[] myStrArray = { "Zero", "One", "Two", "Three" };
Console.WriteLine( String.Join( " ", myStrArray ) );
myStrArray = myStrArray.RemoveAt( 2 );
Console.WriteLine( String.Join( " ", myStrArray ) );
/* Output
* "Zero One Two Three"
* "Zero One Three"
*/
int[] myIntArray = { 0, 1, 2, 3 };
Console.WriteLine( String.Join( " ", myIntArray ) );
myIntArray = myIntArray.RemoveAt( 2 );
Console.WriteLine( String.Join( " ", myIntArray ) );
/* Output
* "0 1 2 3"
* "0 1 3"
*/
}
}
}
Here's how I did it...
public static ElementDefinitionImpl[] RemoveElementDefAt(
ElementDefinition[] oldList,
int removeIndex
)
{
ElementDefinitionImpl[] newElementDefList = new ElementDefinitionImpl[ oldList.Length - 1 ];
int offset = 0;
for ( int index = 0; index < oldList.Length; index++ )
{
ElementDefinitionImpl elementDef = oldList[ index ] as ElementDefinitionImpl;
if ( index == removeIndex )
{
// This is the one we want to remove, so we won't copy it. But
// every subsequent elementDef will by shifted down by one.
offset = -1;
}
else
{
newElementDefList[ index + offset ] = elementDef;
}
}
return newElementDefList;
}
In a normal array you have to shuffle down all the array entries above 2 and then resize it using the Resize method. You might be better off using an ArrayList.
private int[] removeFromArray(int[] array, int id)
{
int difference = 0, currentValue=0;
//get new Array length
for (int i=0; i<array.Length; i++)
{
if (array[i]==id)
{
difference += 1;
}
}
//create new array
int[] newArray = new int[array.Length-difference];
for (int i = 0; i < array.Length; i++ )
{
if (array[i] != id)
{
newArray[currentValue] = array[i];
currentValue += 1;
}
}
return newArray;
}
Here's a small collection of helper methods I produced based on some of the existing answers. It utilizes both extensions and static methods with reference parameters for maximum idealness:
public static class Arr
{
public static int IndexOf<TElement>(this TElement[] Source, TElement Element)
{
for (var i = 0; i < Source.Length; i++)
{
if (Source[i].Equals(Element))
return i;
}
return -1;
}
public static TElement[] Add<TElement>(ref TElement[] Source, params TElement[] Elements)
{
var OldLength = Source.Length;
Array.Resize(ref Source, OldLength + Elements.Length);
for (int j = 0, Count = Elements.Length; j < Count; j++)
Source[OldLength + j] = Elements[j];
return Source;
}
public static TElement[] New<TElement>(params TElement[] Elements)
{
return Elements ?? new TElement[0];
}
public static void Remove<TElement>(ref TElement[] Source, params TElement[] Elements)
{
foreach (var i in Elements)
RemoveAt(ref Source, Source.IndexOf(i));
}
public static void RemoveAt<TElement>(ref TElement[] Source, int Index)
{
var Result = new TElement[Source.Length - 1];
if (Index > 0)
Array.Copy(Source, 0, Result, 0, Index);
if (Index < Source.Length - 1)
Array.Copy(Source, Index + 1, Result, Index, Source.Length - Index - 1);
Source = Result;
}
}
Performance wise, it is decent, but it could probably be improved. Remove relies on IndexOf and a new array is created for each element you wish to remove by calling RemoveAt.
IndexOf is the only extension method as it does not need to return the original array. New accepts multiple elements of some type to produce a new array of said type. All other methods must accept the original array as a reference so there is no need to assign the result afterward as that happens internally already.
I would've defined a Merge method for merging two arrays; however, that can already be accomplished with Add method by passing in an actual array versus multiple, individual elements. Therefore, Add may be used in the following two ways to join two sets of elements:
Arr.Add<string>(ref myArray, "A", "B", "C");
Or
Arr.Add<string>(ref myArray, anotherArray);
I know this article is ten years old and therefore probably dead, but here's what I'd try doing:
Use the IEnumerable.Skip() method, found in System.Linq. It will skip the selected element from the array, and return another copy of the array that only contains everything except the selected object. Then just repeat that for every element you want removed and after that save it to a variable.
For example, if we have an array named "Sample" (of type int[]) with 5 numbers. We want to remove the 2nd one, so trying "Sample.Skip(2);" should return the same array except without the 2nd number.
First step
You need to convert the array into a list, you could write an extension method like this
// Convert An array of string to a list of string
public static List<string> ConnvertArrayToList(this string [] array) {
// DECLARE a list of string and add all element of the array into it
List<string> myList = new List<string>();
foreach( string s in array){
myList.Add(s);
}
return myList;
}
Second step
Write an extension method to convert back the list into an array
// convert a list of string to an array
public static string[] ConvertListToArray(this List<string> list) {
string[] array = new string[list.Capacity];
array = list.Select(i => i.ToString()).ToArray();
return array;
}
Last steps
Write your final method, but remember to remove the element at index before converting back to an array like the code show
public static string[] removeAt(string[] array, int index) {
List<string> myList = array.ConnvertArrayToList();
myList.RemoveAt(index);
return myList.ConvertListToArray();
}
examples codes could be find on my blog, keep tracking.