So I am trying to figure out why filling sudoku board in order of how I generate fields is many times faster then filling it which shuffled order, and I can't come to any reasonable conclusions.
Speed of filling and checking should be about the same no matter if fields are shuffled or not, only noticeable difference is the fact that filling boxes in a shuffled order causes the main loop to easily go for 60.000.000+ iterations not sure how many on average as I am not patient enough to check. And filling them in order of how they were generated gets finished in usually < 1.000 iterations rarely going over it and basically never going over 5.000.
I do understand that code doesn't follow C# standards too strictly but it's a project I am supposed to make and it's not my main language there are probably some possible optimizations but it's not what I am interested in all i want to know why shuffling the order of fill the boxes causes the process to take this much longer.
Edit: forgot to attach order of creating/filling boxes order of filling/creating boxes:
Edit2: more concise rephrased question:
The question is why generating an valid fully filled sudoku board in order from image creates a lot less invalid boards from which there is required backtracking, as compared to generating a board in a random order of choosing fields?
example of random order
Here is the code used for the project https://github.com/Piterm21/sudoku
And here are all the parts used in generation.
filledBox class:
public class filledBox
{
public int textBoxIndex;
public int value;
public int nextIndexToTry;
public int[] values;
public int lineIndex;
public int columnIndex;
public int groupIndex;
}
Here is a main loop:
filledBox[] boxes = this.shuffleBoxesCreateCheckingLists();
long iter = 0;
int nextIndexToFill = 0;
while (nextIndexToFill != 81) {
for (; boxes[nextIndexToFill].nextIndexToTry < 9; boxes[nextIndexToFill].nextIndexToTry++) {
// check if is valid
if (!createsError(boxes[nextIndexToFill])) {
boxes[nextIndexToFill].value = boxes[nextIndexToFill].values[boxes[nextIndexToFill].nextIndexToTry];
nextIndexToFill++;
break;
}
if (boxes[nextIndexToFill].nextIndexToTry == 8) {
boxes[nextIndexToFill].nextIndexToTry = 0;
boxes[nextIndexToFill].value = 0;
nextIndexToFill--;
boxes[nextIndexToFill].nextIndexToTry++;
while (boxes[nextIndexToFill].nextIndexToTry == 9) {
boxes[nextIndexToFill].nextIndexToTry = 0;
boxes[nextIndexToFill].value = 0;
nextIndexToFill--;
boxes[nextIndexToFill].nextIndexToTry++;
}
break;
}
}
iter++;
}
System.Diagnostics.Debug.WriteLine(iter);
Generation of boxes with setting of lists used for checking for errors:
List<filledBox>[] generationLines;
List<filledBox>[] generationColumns;
List<filledBox>[] generationGroups;
public filledBox[] shuffleBoxesCreateCheckingLists()
{
filledBox[] boxes = new filledBox[81];
for (int i = 0; i < 81; i++) {
boxes[i] = new filledBox();
boxes[i].values = new int[9]{ 1, 2, 3, 4, 5, 6, 7, 8, 9 };
boxes[i].values = shuffle(boxes[i].values, true);
}
generationLines = new List<filledBox>[9];
generationColumns = new List<filledBox>[9];
generationGroups = new List<filledBox>[9];
for (int i = 0; i < 9; i++) {
generationLines[i] = new List<filledBox>();
generationColumns[i] = new List<filledBox>();
generationGroups[i] = new List<filledBox>();
}
int boxIndex = 0;
for (int y = 0; y < 3; y++) {
for (int x = 0; x < 3; x++) {
for (int innerY = 0; innerY < 3; innerY++) {
for (int innerX = 0; innerX < 3; innerX++) {
int subPanelIndex = x + y * 3;
int lineIndex = innerY + y * 3;
int columnIndex = innerX + x * 3;
boxes[boxIndex].textBoxIndex = boxIndex;
boxes[boxIndex].groupIndex = subPanelIndex;
boxes[boxIndex].columnIndex = columnIndex;
boxes[boxIndex].lineIndex = lineIndex;
boxes[boxIndex].nextIndexToTry = 0;
boxes[boxIndex].value = 0;
generationLines[lineIndex].Add(boxes[boxIndex]);
generationColumns[columnIndex].Add(boxes[boxIndex]);
generationGroups[subPanelIndex].Add(boxes[boxIndex]);
boxIndex++;
}
}
}
}
#if !fast
boxes = shuffle(boxes);
#endif
return boxes;
}
Shuffling code:
public T[] shuffle<T> (T[] array)
{
int i = array.Length;
while (i > 1) {
i--;
int j = rnd.Next(0, i - 1);
T temp = array[i];
array[i] = array[j];
array[j] = temp;
}
return array;
}
And code responsible for checking for errors:
public bool hasError (List<filledBox> list, filledBox box)
{
bool result = false;
foreach (filledBox filledBoxToCheck in list) {
if (filledBoxToCheck.value == box.values[box.nextIndexToTry]) {
if (box.lineIndex != filledBoxToCheck.lineIndex ||
box.columnIndex != filledBoxToCheck.columnIndex) {
result = true;
break;
}
}
}
return result;
}
public bool createsError (filledBox box)
{
bool result = false;
if (hasError(generationGroups[box.groupIndex], box)) {
result = true;
} else if (hasError(generationLines[box.lineIndex], box)) {
result = true;
} else if (hasError(generationColumns[box.columnIndex], box)) {
result = true;
}
return result;
}
The reason the shuffled order of visiting boxes is worse is that, for the first few boxes, you've increased the odds of filling them in independently. I'm calling a 3x3 region a box. You could visit each little square randomly, but I'll confine my analysis to fillng in one 3x3 box at a time. It should illustrate the problem.
Lets assume, whether you used the in-order or random method, you filled in the top left box first.
VISITING BOXES IN ORDER:
Now consider filling in the box next to it, that is, the top middle box. Just look at filling in the top row of that box, there are 6x5x4 = 120 ways to fill it in.
VISITING BOXES RANDOMLY:
Now consider instead choosing any box to fill in next. If you happen to choose a box that is not in the top row or left column, filling it in is unaffected by the way the first box was filled. The top row of that box can be filled in 9x8x7 = 504 ways.
Taking this further, if you fill in the boxes sequentially, say, across the top, the third box (the top right one), begins with only 3x2x1 = 6 ways to fill in the top row of it. Whereas, if you select the next box randomly, you might, at worst, select a box that is not in the same row or column of either of the first 2 boxes, which means that the top row of that box has yet again 9x8x7 = 504 ways of being filled in.
If you tried randomly selecting the order for each little square to be filled in, the first few might all be in different rows and columns. At worst, none of the first 9 squares will align meaning there will be 9^9 = 387,420,489 options. But filling them in across the top row, the choices are guaranteed to be 9! = 362,880.
This illustrates one of the strategies for how to approach backtracking. Ideally, you first solve parts of the problem that most tightly constrain the rest of the problem. For some problems sequential is better, and for some random is better.
Related
I'm trying to create some pairs of unique numbers using pretty simple algorithm.
For some unknown reason after compiling Unity goes into an endless "not responding" state. Seems like it's stuck in a do..while loop, but I don't see any reason for that.
//Creating two lists to store random numbers
List<int> xList = new List<int>();
List<int> yList = new List<int>();
int rx, ry;
for(int i = 0; i < 10; i++)
{
// look for numbers until they are unique(while they are in lists)
do
{
rx = rand.Next(0, width);
ry = rand.Next(0, height);
}
while(xList.Contains(rx) || yList.Contains(ry));
//add them to lists
xList.Add(rx);
yList.Add(ry);
Debug.Log(rx + ", " + ry);
// some actions with these numbers
gridArray[rx,ry].isBomb = true;
gridArray[rx,ry].changeSprite(bombSprite);
}
As mentioned the issue is that once all unique numbers have been used once you are stuck in the do - while loop.
Instead you should rather simply
generate the plain index lists for all possible pairs.
I will use the Unit built-in type Vector2Int but you could do the same using your own struct/class
For each bomb to place pick a random entry from the list of pairs
Remove according random picked item from the pairs so it is not available anymore in the next go
Something like
// create the plain pair list
var pairs = new List<Vector2Int>(width * height);
for(var x = 0; x < width; x++)
{
for(var y = 0; y < height; y++)
{
pairs.Add(new Vector2Int(x,y));
}
}
// so now you have all possible permutations in one list
if(pairs.Count < BOMB_AMOUNT_TO_PLACE)
{
Debug.LogError("You are trying more bombs than there are fields in the grid!");
return;
}
// Now place your bombs one by one on a random spot in the grid
for(var i = 0; i < BOMB_AMOUNT_TO_PLACE; i++)
{
// now all you need to do is pick one random index from the possible entries
var randomIndexInPairs = Random.Range(0, pairs.Count);
var randomPair = pairs[randomIndexInPairs];
// and at the same time remove the according entry
pairs.RemoveAt(randomIndexInPairs);
// Now you have completely unique but random index pairs
var rx = randomPair.x;
var ry = randomPair.y;
gridArray[rx, ry].isBomb = true;
gridArray[rx, ry].changeSprite(bombSprite);
}
Depending on your use-case as alternative to generate the pairs list and then remove entries again you could also generate it once and then use
if(pairs.Count < BOMB_AMOUNT_TO_PLACE)
{
Debug.LogError("You are trying more bombs than there are fields in the grid!");
return;
}
var random = new System.Random();
var shuffledPairs = pairs.OrderBy(e => random.Next());
for(var i = 0; i < BOMB_AMOUNT_TO_PLACE; i++)
{
// then you can directly use
var randomPair = shuffledPairs[i];
// Now you have completely unique but random index pairs
var rx = randomPair.x;
var ry = randomPair.y;
gridArray[rx, ry].isBomb = true;
gridArray[rx, ry].changeSprite(bombSprite);
}
Although your algorithm is maybe not the best way to generate ten bombs in a grid, it should work.
The problem is that your while condition is using a OR statement, which means that if you have a bomb in the first line (in any column), it will not be able to add another bomb in that line.
Therefore you will pretty soon end up with an infinite loop because for every bomb you lock the line and column.
If you put an AND condition, you make sure the pair is unique because you lock only that cell.
Provided of course that width x height is more than ten.
I have code that randomly selects 15 sprites from 51 sprites. It can generate 2 duplicated sprites (this is acceptable for me) but I want to avoid 3 duplicated values. How can I prevent that?
My code
for (int i = 0; i < 16; i++)
{
int arrIndex = UnityEngine.Random.Range(0, tasSprites.Length);
tasSprite = tasSprites[arrIndex];
tasName = tasSprite.name;
taslar.Add(Int32.Parse(tasName));
}
Have you tried checking if each generated index already appears twice in taslar and if so, generating another one?
while (taslar.Count < 16)
{
int arrIndex = UnityEngine.Random.Range(0, tasSprites.Length);
tasSprite = tasSprites[arrIndex];
tasName = tasSprite.name;
int value = Int32.Parse(tasName);
if (taslar.Count(t => t == value) < 2)
{
taslar.Add(value);
}
}
You could do something like:
int[] indexesCount = new int[tasSprites.Length];
while (taslar.Count < 16)
{
int arrIndex = UnityEngine.Random.Range(0, tasSprites.Length);
if (indexesCount[arrIndex] == 2)
{
continue;
}
indexesCount[arrIndex]++;
tasSprite = tasSprites[arrIndex];
tasName = tasSprite.name;
taslar.Add(Int32.Parse(tasName));
}
Note that this solution is "good" while tasSprites.Length is relatively small. We are creating a temporary array of size tasSprites.Length to see which numbers have already been used.
Generate an array with each of your sprites in twice
Select 15 values from the array using e.g. a Fisher-Yates shuffle
This will save you the chance of possibly repeatedly generating values that have already been used twice.
(this is a library)
The function GetUniqueInt is being called with (5, 5) as the variables.
Currently the code will stall unity to a complete halt, or crash my PC with a memory overflow error.
Does anyone have any ideas as to how I could prevent it from crashing or what is making it crash?
using UnityEngine;
namespace MajorSolution
{
public static class MajorMath
{
public static int[] GetUniqueInt(int intCount, int intLength)
{
int[] returnValue = new int[intCount];
int[] temp = new int[intLength];
for (int a = 0; a < intCount; a++)
{
string create = new string("create".ToCharArray());
switch (create)
{
case "create":
returnValue[a] = GetRandomInt(intCount);
goto case "check";
case "check":
bool alreadyTaken = false;
for (int c = 0; c < returnValue.Length - 1; c++)
{
if (returnValue[a] == returnValue[c])
{
// Already Taken!
alreadyTaken = true;
}
}
if (!alreadyTaken)
{
break;
}
else
{
goto case "create";
}
}
}
Debug.Log(returnValue);
return returnValue;
}
public static int GetRandomInt(int intCount)
{
int[] storage = new int[intCount];
int returnValue = 0;
for (int i = 0; i < intCount; i++)
{
storage[i] = (Mathf.FloorToInt(Random.Range(0, 9)) * (int)Mathf.Pow(10,i));
returnValue += storage[i];
}
return returnValue;
}
}
}
Edit I just realized I did not exactly answer the question of why it is bringing the PC to a halt because you have an infinite loop in the code.
The problem is occurring in the following lines of code, notice what is happening.
case "create":
returnValue[a] = GetRandomInt(intCount);
goto case "check";
In the above block of code you are generating a number and putting it in the returnValue array. Now you jump into your "check" block
case "check":
bool alreadyTaken = false;
for (int c = 0; c < returnValue.Length - 1; c++)
{
if (returnValue[a] == returnValue[c])
{
// Already Taken!
alreadyTaken = true;
}
}
In this block of code you are looping over the entire returnValue array including the value you just inserted in it. Basically you are looping over the array asking if a value that you just put in the array is in the array.
Without knowing exactly what you are trying to do with these methods, I will just make a simple fix suggestion with some minor cleanup
public static int[] GetUniqueInt(int count, int length)
{
var returnValue = new int[count];
var values = new HashSet<int>(); // Used to track what numbers we have generated
for (int i = 0; i < count; ++i)
{
// Generate the number and check to be sure we haven't seen it yet
var number = GetRandomInt(length);
while(values.Contains(number)) // This checks if the number we just generated exists in the HashSet of seen numbers
{
// We get here if the HashSet contains the number. If we have
// seen the number then we need to generate a different one
number = GetRandomInt(length);
}
// When we reach this point, it means that we have generated a new unique number
// Add the number to the return array and also add it to the list of seen numbers
returnValue[a] = number;
values.Add(number); // Adds the number to the HashSet
}
Debug.Log(returnValue);
return returnValue;
}
I did end up removing the using of intLength but from your posted code it was only used to declare a temp array which itself was never used. Based on that, I just removed it entirely.
Based on your comment I updated the fix to use intLength. I made one other minor change. I removed int from the variable names of count and length. Hungarian notation is a lot less common in C# code. Personally, I feel like the code is cleaner and easier to read without the Hungarian notation. The key is to use good variable names that express the intent or make it easier to follow. In this case count is the count (read total number) of numbers you want returned and length is the length of the number. You could consider maybe even renaming that to numberOfDigits to make it clearer that the idea is you are going to create a random number with that number of digits in it.
In my current project I have an algorythm that generats data, this data will be shown in a Linechart, however the chart only updates once the algorythm has ended rather than after each "step".
private void simulate(share[] shares)
{
int k = 0;
Random r = new Random(); //random values just for testing
while (k < 10)//10 steps (10 values for each share)
{
for (int i = 0; i < shares.Length; i++)
{
shares[i].value = r.Next(0, 10000);//random a new value
shares[i].history.Add(shares[i].value);//add the value to the value history
}
//now update the chart so that it first shows only one x point than two and so on
drawchart(shares);
k++;
}
}
private void drawchart(share[] shares)
{
cHshares.Series.Clear();//clear the chart
for (int i = 0; i < shares.Length; i++)//for each share
{
cHshares.Series.Add(shares[i].name);//add a new share to the chart
cHshares.Series[shares[i].name].ChartType = SeriesChartType.FastLine;
int j = 0;
//draw the lines with each value that exists for each share
foreach (double value in shares[i].history)
{
cHshares.Series[shares[i].name].Points.AddXY(j, value);
j++;
}
}
}
Since I call the drawchart funktion every step why is it that it only shows after all steps are finished?
you have put your code in a loop like for loop
After each drawchart, update your chart. i.e.
drawchart(shares);
cHshares.Update();
k++;
I hope this will be useful.
Hi there I'm trying to write a method that reads every number in a list and detects where it spikes and drops. This is what I have so far:
I basically figure if I loop through the list, loop through it again to get the next number in the list, then detecting if it's more or less. If it's more it'll save to one list, vice versa.
What I want this method to do is determine where there's a spike of 100 or more, save the point that it does this (which is 'counter') and also save the points where the numbers drop.
This so far notices only a drop and it will save every number in the list until it spikes again and once it has spiked it shows no numbers, until it drops again and so on.
I've put 'check' and 'check2' to try and counteract it saving every number after it notices a drop and only save it once but no luck.
Any ideas?
public void intervalDetection()
{
//Counter is the point in the list
int counter = 0;
int spike = 0;
int drop = 0;
//Loop through power list
for (int i = 0; i < powerList.Count(); i++)
{
counter++;
int firstNumber = powerList[i];
//Loop again to get the number after??
for (int j = 1; j < 2; j++)
{
//Detect Spike
spike = firstNumber + 100;
drop = firstNumber - 100;
if (powerList[j] > spike)
{
if (check2 == false)
{
intervalStartList.Add(counter);
check2 = true;
check = false;
}
}
//Detect Drop
else if (powerList[j] < drop)
{
if (check == false)
{
intervalEndList.Add(counter);
check = true;
check2 = false;
}
}
}
Create integer "average"
Loop through List/Array and add each value to average
Divide average by the count of the List/Array
Loop through List/Array and check deviation to the average integer
derp
Code example:
public class DSDetector {
public static List<int>[] getDropsnSpikes(List<int> values, int deviation) {
List<int> drops = new List<int>();
List<int> spikes = new List<int>();
int average = 0;
foreach (int val in values) {
average += val;
}
average = average/values.Count;
foreach (int val in values) {
if (val < average - deviation) {
drops.add(val);
}
if (val > average + deviation) {
spikes.add(val);
}
}
//derp.
return new List<int>{drops, spikes};
}
}
not tested but I think it works. Just try it.
What exactly do you mean saying "peaks" and "drops"?
Let's say you have following list of integers
112, 111, 113, 250, 112, 111, 1, 113
In this case value 250 is peak and 1 drop relative to average value and you can get it using Kai_Jan_57 answer.
But also 250 is peak to previous value 113 and 112 is drop for 250.
If you want to find local peaks and drops you can check each value relative to previous and next: find average as avg=(val[i-1]+val[i+1])/2 and check if val[i]>avg + 100 (peak) or val[i]