squidpony.squidgrid.DetailedMimic Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of squidlib-util Show documentation
Show all versions of squidlib-util Show documentation
SquidLib platform-independent logic and utility code. Please refer to
https://github.com/SquidPony/SquidLib .
/*
The MIT License(MIT)
Copyright(c) mxgmn 2016.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
The software is provided "as is", without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose and noninfringement. In no event shall the authors or copyright holders be liable for any claim, damages or other liability, whether in an action of contract, tort or otherwise, arising from, out of or in connection with the software or the use or other dealings in the software.
*/
package squidpony.squidgrid;
import squidpony.ArrayTools;
import squidpony.squidmath.*;
/**
* Similar to MimicFill, this class can be used to imitate the style of an existing piece of data, but this works on
* more than just booleans; it can produce similar styles of texture (its original use in SynTex), of map, of item
* placement, and so on by specifying a different technique for differentiating two int values.
*
* Two options are now available; the process() method allows the slow analyze() step to be computed once,
* before the rest of processing is potentially done many times, but the new neoProcess() method produces
* comparable or better results and is drastically faster even without needing analysis. This means
* neoProcess(), based on P.F. Harrison's algorithm in SynTex, is strongly recommended now.
*
* Example results of neoProcess(), with a procedural dungeon first and the mimic versions after it:
* https://gist.github.com/tommyettinger/405336fb0fc74838d806c021aabe77da (you may need to click Raw and zoom
* out somewhat for it to render well).
*
* Ported from https://github.com/mxgmn/SynTex by Tommy Ettinger on 6/9/2016.
*/
public class DetailedMimic {
/**
* Constructor that uses an unseeded RNG and, without any instruction otherwise, assumes the ints this is asked to
* compare are colors in RGBA8888 format. You can specify your own implementation of the AestheticDifference
* interface (one function) and pass it to other constructors, as well.
*/
public DetailedMimic()
{
this(AestheticDifference.rgba8888);
}
/**
* Constructor that uses an unseeded RNG (effectively a random seed) and the given AestheticDifference. An example
* piece of code that implements an AestheticDifference is available in the docs for
* {@link AestheticDifference#difference(int, int)}; it is also considered a functional interface if you use Java 8
* or newer. You can also use the ready-made implementation {@link AestheticDifference#rgba8888} if you have int
* data that represents RGBA8888 colors, which can be obtained from libGDX Colors or SColors in the display module.
* @param diff an implementation of the AestheticDifference interface, such as {@link AestheticDifference#rgba8888};
* may be null, but that forces all calls to processing methods to treat discrete as true
*/
public DetailedMimic(AestheticDifference diff)
{
this(diff, new GWTRNG());
}
/**
* Constructor that uses the given RNG and the given AestheticDifference. An example
* piece of code that implements an AestheticDifference is available in the docs for
* {@link AestheticDifference#difference(int, int)}; it is also considered a functional interface if you use Java 8
* or newer. You can also use the ready-made implementation {@link AestheticDifference#rgba8888} if you have int
* data that represents RGBA8888 colors, which can be obtained from libGDX Colors or SColors in the display module.
* @param diff an implementation of the AestheticDifference interface, such as {@link AestheticDifference#rgba8888};
* may be null, but that forces all calls to processing methods to treat discrete as true
* @param rng an IRNG, such as an RNG, to generate random factors; may be seeded to produce reliable output
*/
public DetailedMimic(AestheticDifference diff, IRNG rng)
{
random = rng;
difference = diff;
analyzed = null;
}
/**
* DISCOURAGED; use {@link #neoProcess(int[], int, int, int, int, int, int, boolean)} instead, which doesn't need a
* separate analysis step.
* Analyzes a sample as a 1D int array and stores the needed info to call
* {@link #process(int[], int, int, int, int, int, int, double, boolean)} any number of times later on without
* recalculating some heavy-weight information.
* @param sample a 1D array of ints that can be compared by the AestheticDifference this uses (or any ints if
* discrete is true)
* @param width the width of the area in sample this should use (sample can be interpreted as different shapes)
* @param height the height of the area in sample this should use (sample can be interpreted as different shapes)
* @param detailLevel how much detail to try for; if 0 or less this does nothing, 2 works well in general
* @param proximity how far away to consider cells as affecting another; 3 works well
* @param discrete false if this can produce ints other than those in the input; true if it uses a fixed set
*/
public void analyze(int[] sample, int width, int height, int detailLevel, int proximity, boolean discrete)
{
discrete |= (difference == null);
if(detailLevel > 0)
analysis(sample, width, height, detailLevel, proximity, discrete);
}
/**
* DISCOURAGED; use {@link #neoProcess(int[], int, int, int, int, int, int, boolean)} instead, which doesn't need a
* separate analysis step.
* Processes a sample as a 1D int array and returns a different 1D int array that mimics the input. If the last time
* this was called used the same sample, sampleWidth, and sampleHeight parameters, or if
* {@link #analyze(int[], int, int, int, int, boolean)} was called with its width equal to sampleWidth and its
* height equal to sampleHeight, then this doesn't need to perform as many expensive calculations.
* @param sample a 1D array of ints that can be compared by the AestheticDifference this uses (or any ints if
* discrete is true)
* @param sampleWidth the width of the area in sample this should use (sample can be interpreted as different shapes)
* @param sampleHeight the height of the area in sample this should use (sample can be interpreted as different shapes)
* @param targetWidth the desired width of the output
* @param targetHeight the desired height of the output
* @param detailLevel how much detail to try for; if 0 or less this doesn't perform analysis and has somewhat lower
* quality, but 2 works well in general
* @param proximity how far away to consider cells as affecting another; 3 works well
* @param temperature a level of unpredictability in the output relative to the input; must be greater than 0
* @param discrete false if this can produce ints other than those in the input; true if it uses a fixed set
* @return a new 1D int array that can be interpreted as having targetWidth and targetHeight, and mimics sample
*/
public int[] process(int[] sample, int sampleWidth, int sampleHeight, int targetWidth, int targetHeight,
int detailLevel, int proximity, double temperature, boolean discrete)
{
discrete |= (difference == null);
if(detailLevel > 0)
{
if(analyzed == null || analyzed.length != sampleWidth || analyzed.length == 0 ||
analyzed[0].length != sampleHeight)
analyze(sample, sampleWidth, sampleHeight, detailLevel, proximity, discrete);
return coherentSynthesis(sample, sampleWidth, sampleHeight, analyzed, proximity,
targetWidth, targetHeight, temperature, discrete);
}
else {
return fullSynthesis(sample, sampleWidth, sampleHeight, proximity,
targetWidth, targetHeight, temperature, discrete);
}
}
/**
* Processes a 1D int array representing 2D storage of values that can be compared by this object's
* AestheticDifference (or any values if that is null or discrete is true), and returns a 1D array representing data
* with potentially different dimensions but similar appearance to sample.
* @param sample a 1D array of ints that can be compared by the AestheticDifference this uses (or any ints if
* discrete is true)
* @param sampleWidth the width of the area in sample this should use (sample can be interpreted as different shapes)
* @param sampleHeight the height of the area in sample this should use (sample can be interpreted as different shapes)
* @param targetWidth the desired width of the output
* @param targetHeight the desired height of the output
* @param detailLevel how much detail to try for; here this will always be treated as at least 1
* @param proximity how far away to consider cells as affecting another; 3 works well
* @param discrete false if this can produce ints other than those in the input; true if it uses a fixed set
* @return a new 1D int array that can be interpreted as having targetWidth and targetHeight, and mimics sample
*/
public int[] neoProcess(int[] sample, int sampleWidth, int sampleHeight, int targetWidth, int targetHeight,
int detailLevel, int proximity, boolean discrete)
{
return reSynthesis(sample, sampleWidth, sampleHeight, proximity, 20,
Math.max(1, detailLevel), discrete || (difference == null), targetWidth, targetHeight);
}
public IRNG random;
public AestheticDifference difference;
private int[][] analyzed;
private void analysis(int[] bitmap, int width, int height, int K, int N, boolean indexed)
{
int area = width * height;
analyzed = new int[area][];
OrderedSet points = new OrderedSet<>(area);
for (int i = 0; i < area; i++) points.add(i);
double[] similarities = new double[area * area];
for (int i = 0; i < area; i++) for (int j = 0; j < area; j++)
similarities[i * area + j] = similarities[j * area + i] != 0 ? similarities[j * area + i] :
similarity(i, bitmap, width, height, j, bitmap, width, height, N, null, indexed);
for (int i = 0; i < area; i++)
{
analyzed[i] = new int[K];
OrderedSet copy = new OrderedSet<>(points);
analyzed[i][0] = i;
copy.remove(i);
for (int k = 1; k < K; k++)
{
double max = -10000;
int argmax = -1;
for(Integer p : copy)
{
double s = similarities[i * area + p];
if (s > max)
{
max = s;
argmax = p;
}
}
analyzed[i][k] = argmax;
copy.remove(argmax);
}
}
}
private int[] coherentSynthesis(int[] sample, int SW, int SH, int[][] sets, int N, int OW, int OH, double t, boolean indexed)
{
int[] result = new int[OW * OH];
Integer[] origins = new Integer[OW * OH];
boolean[][] mask = new boolean[SW][SH], cleanMask = new boolean[SW][SH];
for (int i = 0; i < OW * OH; i++)
{
int x = i % OW, y = i / OW;
IntDoubleOrderedMap candidates = new IntDoubleOrderedMap();
ArrayTools.insert(cleanMask, mask, 0, 0);
for (int dy = -1; dy <= 1; dy++){
for (int dx = -1; dx <= 1; dx++)
{
int sx = (x + dx + OW) % OW, sy = (y + dy + OH) % OH;
Integer origin = origins[sy * OW + sx];
if ((dx != 0 || dy != 0) && origin != null)
{
for (int pi = 0, p; pi < sets[origin].length; pi++)
{
p = sets[origin][pi];
int ox = (p % SW - dx + SW) % SW, oy = (p / SW - dy + SH) % SH;
double s = similarity(oy * SW + ox, sample, SW, SH, i, result, OW, OH, N, origins, indexed);
if (!mask[ox][oy]) candidates.put(ox + oy * SW, Math.pow(100, s / t));
mask[ox][oy] = true;
}
}
}
}
int shifted = candidates.isEmpty() ? random.nextInt(SW) + random.nextInt(SH) * SW : weightedRandom(candidates, random.nextDouble());
origins[i] = shifted;
result[i] = sample[shifted];
}
return result;
}
private int[] fullSynthesis(int[] sample, int SW, int SH, int N, int OW, int OH, double t, boolean indexed)
{
int[] result = new int[OW * OH];
Integer[] origins = new Integer[OW * OH];
if (!indexed) for (int y = 0; y < OH; y++) {
for (int x = 0; x < OW; x++){
if (y + N >= OH)
{
result[x + y * OW] = sample[random.nextInt(SW * SH)];
origins[x + y * OW] = -1;
}
}
}
for (int i = 0; i < OW * OH; i++)
{
double[] candidates = new double[SW * SH];
double max = -10000;
int argmax = -1;
for (int j = 0; j < SW * SH; j++)
{
double s = similarity(j, sample, SW, SH, i, result, OW, OH, N, origins, indexed);
if (s > max)
{
max = s;
argmax = j;
}
if (indexed) candidates[j] = Math.pow(100.0, s / t);
}
if (indexed) argmax = weightedRandom(candidates, random.nextDouble());
result[i] = sample[argmax];
origins[i] = -1;
}
return result;
}
private int[] reSynthesis(int[] sample, int SW, int SH, int N, int M, int polish, boolean indexed, int OW, int OH)
{
IntVLA colors = new IntVLA();
int[] indexedSample = new int[Math.min(SW * SH, sample.length)];
for (int j = 0; j < indexedSample.length; j++)
{
int color = sample[j];
int i = 0;
for (int cn = 0; cn < colors.size; cn++)
{
if(colors.get(cn) == color) break;
i++;
}
if (i == colors.size) colors.add(color);
indexedSample[j] = i;
}
int colorsNumber = colors.size;
double[][] colorMetric = null;
if (!indexed && colorsNumber <= 1024)
{
colorMetric = new double[colorsNumber][colorsNumber];
for (int x = 0; x < colorsNumber; x++)
{
for (int y = 0; y < colorsNumber; y++)
{
int cx = colors.get(x), cy = colors.get(y);
colorMetric[x][y] = difference.difference(cx, cy);
}
}
}
int[] origins = new int[OW * OH];
for (int i = 0; i < origins.length; i++) origins[i] = -1;
int[] shuffle = new int[OW * OH];
for (int i = 0; i < shuffle.length; i++)
{
int j = random.nextInt(i + 1);
if (j != i) shuffle[i] = shuffle[j];
shuffle[j] = i;
}
for (int round = 0; round <= polish; round++) for (int counter = 0; counter < shuffle.length; counter++)
{
int f = shuffle[counter];
int fx = f % OW, fy = f / OW;
int neighborsNumber = round > 0 ? 8 : Math.min(8, counter);
int neighborsFound = 0;
int[] candidates = new int[neighborsNumber + M];
if (neighborsNumber > 0)
{
int[] neighbors = new int[neighborsNumber];
int[] x = new int[4], y = new int[4];
for (int radius = 1; neighborsFound < neighborsNumber; radius++)
{
x[0] = fx - radius;
y[0] = fy - radius;
x[1] = fx - radius;
y[1] = fy + radius;
x[2] = fx + radius;
y[2] = fy + radius;
x[3] = fx + radius;
y[3] = fy - radius;
for (int k = 0; k < 2 * radius; k++)
{
for (int d = 0; d < 4; d++)
{
x[d] = (x[d] + 10 * OW) % OW;
y[d] = (y[d] + 10 * OH) % OH;
if (neighborsFound >= neighborsNumber) continue;
int point = x[d] + y[d] * OW;
if (origins[point] != -1)
{
neighbors[neighborsFound] = point;
neighborsFound++;
}
}
y[0]++;
x[1]++;
y[2]--;
x[3]--;
}
}
for (int n = 0; n < neighborsNumber; n++)
{
int cx = (origins[neighbors[n]] + (f - neighbors[n]) % OW + 100 * SW) % SW;
int cy = (origins[neighbors[n]] / SW + f / OW - neighbors[n] / OW + 100 * SH) % SH;
candidates[n] = cx + cy * SW;
}
}
for (int m = 0; m < M; m++) candidates[neighborsNumber + m] = random.nextInt(SW * SH);
double max = -1E+10;
int argmax = -1;
for (int c = 0; c < candidates.length; c++)
{
double sum = random.nextDouble(0.000001);
int ix = candidates[c] % SW, iy = candidates[c] / SW, jx = f % OW, jy = f / OW;
int SX, SY, FX, FY, S, F;
int origin;
for (int dy = -N; dy <= N; dy++)
{
for (int dx = -N; dx <= N; dx++)
{
if (dx != 0 || dy != 0)
{
SX = ix + dx;
if (SX < 0) SX += SW;
else if (SX >= SW) SX -= SW;
SY = iy + dy;
if (SY < 0) SY += SH;
else if (SY >= SH) SY -= SH;
FX = jx + dx;
if (FX < 0) FX += OW;
else if (FX >= OW) FX -= OW;
FY = jy + dy;
if (FY < 0) FY += OH;
else if (FY >= OH) FY -= OH;
S = SX + SY * SW;
F = FX + FY * OW;
origin = origins[F];
if (origin != -1)
{
if (indexed) sum += sample[origin] == sample[S] ? 1 : -1;
else if (colorMetric != null) sum += colorMetric[indexedSample[origin]][indexedSample[S]];
else sum += difference.difference(sample[origin], sample[S]);
}
}
}
}
if (sum >= max)
{
max = sum;
argmax = candidates[c];
}
}
origins[f] = argmax;
}
int[] result = new int[OW * OH];
for (int i = 0; i < result.length; i++) result[i] = sample[origins[i]];
return result;
}
private double similarity(int i1, int[] b1, int w1, int h1, int i2, int[] b2, int w2, int h2, int N, Integer[] origins, boolean indexed)
{
double sum = 0;
int x1 = i1 % w1, y1 = i1 / w1, x2 = i2 % w2, y2 = i2 / w2;
for (int dy = -N; dy <= 0; dy++) for (int dx = -N; (dy < 0 && dx <= N) || (dy == 0 && dx < 0); dx++)
{
int sx1 = (x1 + dx + w1) % w1, sy1 = (y1 + dy + h1) % h1;
int sx2 = (x2 + dx + w2) % w2, sy2 = (y2 + dy + h2) % h2;
int c1 = b1[sx1 + sy1 * w1];
int c2 = b2[sx2 + sy2 * w2];
if (origins == null || origins[sy2 * w2 + sx2] != null)
{
if (indexed)
sum += c1 == c2 ? 1 : -1;
else
sum -= difference.difference(c1, c2);
/*
Color C1 = Color.FromArgb(c1), C2 = Color.FromArgb(c2);
sum -= (double)((C1.R - C2.R) * (C1.R - C2.R) + (C1.G - C2.G) * (C1.G - C2.G) + (C1.B - C2.B) * (C1.B - C2.B)) / 65536.0;
*/
}
}
return sum;
}
static int weightedRandom(double[] array, double r)
{
double sum = 0;
for (int j = 0; j < array.length; j++)
sum += Math.max(1.0, array[j]);
for (int j = 0; j < array.length; j++)
array[j] /= sum;
int i = 0;
double x = 0;
while (i < array.length)
{
x += array[i];
if (r <= x) return i;
i++;
}
return 0;
}
static int weightedRandom(IntDoubleOrderedMap dic, double r) {
// int[] ints = dic.keySet().toIntArray();
// double[] doubles = dic.values().toDoubleArray();
double sum = 0;
final int sz = dic.size();
for (int j = 0; j < sz; j++)
sum += Math.max(1.0, dic.getAt(j));
int i = 0;
double x = 0;
while (i < sz)
{
x += dic.getAt(i) / sz;
if (r <= x) return dic.keyAt(i);
i++;
}
return dic.firstKey();
}
/**
* Utility method to produce 1D int arrays this can process when discrete is true or difference is null.
* @param map a 2D char array
* @return an int array that can be used as a sample
*/
public static int[] convertCharToInt(char[][] map)
{
int w = map.length, h = map[0].length;
int[] result = new int[w * h];
for (int x = 0; x < w; x++) {
for (int y = 0; y < h; y++) {
result[x * h + y] = map[x][y];
}
}
return result;
}
/**
* Utility method that takes a 1D int array that represents chars (such as a sample produced by
* {@link #convertCharToInt(char[][])} or, more likely, the result of processing such a sample with this class) and
* returns a 2D char array with the requested width and height (which should match the targetWidth and targetHeight
* given during processing).
* @param arr a 1D int array that represents char values
* @param w the width that arr can be interpreted as; should probably match the targetWidth given in processing
* @param h the height that arr can be interpreted as; should probably match the targetHeight given in processing
* @return a 2D char array with the given width and height, probably filled with the data from arr
*/
public static char[][] convertIntToChar(int[] arr, int w, int h)
{
char[][] result = new char[w][h];
if(arr == null)
return result;
for (int x = 0; x < w; x++) {
for (int y = 0; y < h; y++) {
if(x * h + y >= arr.length)
return result;
result[x][y] = (char) arr[x * h + y];
}
}
return result;
}
}