private static Vec3 GetMultiplicity1Evector(Sym3x3 matrix, float evalue)
{
//
// compute M
//
Sym3x3 m = new Sym3x3
{
[0] = matrix[0] - evalue,
[1] = matrix[1],
[2] = matrix[2],
[3] = matrix[3] - evalue,
[4] = matrix[4],
[5] = matrix[5] - evalue
};
//
// compute U
//
Sym3x3 u = new Sym3x3
{
[0] = m[3] * m[5] - m[4] * m[4],
[1] = m[2] * m[4] - m[1] * m[5],
[2] = m[1] * m[4] - m[2] * m[3],
[3] = m[0] * m[5] - m[2] * m[2],
[4] = m[1] * m[2] - m[4] * m[0],
[5] = m[0] * m[3] - m[1] * m[1]
};
//
// find the largest component
//
float mc = Math.Abs(u[0]);
int mi = 0;
for (int i = 1; i < 6; ++i)
{
float c = Math.Abs(u[i]);
if (c > mc)
{
mc = c;
mi = i;
}
}
//
// pick the column with this component
//
switch (mi)
{
case 0: return(new Vec3(u[0], u[1], u[2]));
case 1:
case 3: return(new Vec3(u[1], u[3], u[4]));
default: return(new Vec3(u[2], u[4], u[5]));
}
}
private static Vec3 GetMultiplicity2Evector(Sym3x3 matrix, float evalue)
{
//
// compute M
//
Sym3x3 m = new Sym3x3
{
[0] = matrix[0] - evalue,
[1] = matrix[1],
[2] = matrix[2],
[3] = matrix[3] - evalue,
[4] = matrix[4],
[5] = matrix[5] - evalue
};
//
// find the largest component
//
float mc = Math.Abs(m[0]);
int mi = 0;
for (int i = 1; i < 6; ++i)
{
float c = Math.Abs(m[i]);
if (c > mc)
{
mc = c;
mi = i;
}
}
//
// pick the first eigenvector based on this index
//
switch (mi)
{
case 0:
case 1: return(new Vec3(-m[1], m[0], 0.0f));
case 2: return(new Vec3(m[2], 0.0f, -m[0]));
case 3:
case 4: return(new Vec3(0.0f, -m[4], m[3]));
default: return(new Vec3(0.0f, -m[5], m[4]));
}
}
private const float FLT_EPSILON = 1.192092896e-07F; // This is not the same value as Single.Epsilon
#endregion Private Fields
#region Public Methods
public static Sym3x3 ComputeWeightedCovariance(int n, Vec3[] points, float[] weights)
{
//
// compute the centroid
//
float total = 0.0f;
Vec3 centroid = new Vec3(0.0f);
for (int i = 0; i < n; ++i)
{
total += weights[i];
centroid += weights[i] * points[i];
}
centroid /= total;
//
// accumulate the covariance matrix
//
Sym3x3 covariance = new Sym3x3(0.0f);
for (int i = 0; i < n; ++i)
{
Vec3 a = points[i] - centroid;
Vec3 b = weights[i] * a;
covariance[0] += a.X * b.X;
covariance[1] += a.X * b.Y;
covariance[2] += a.X * b.Z;
covariance[3] += a.Y * b.Y;
covariance[4] += a.Y * b.Z;
covariance[5] += a.Z * b.Z;
}
//
// return it
//
return(covariance);
}
private const float FLT_EPSILON = 1.192092896e-07F; // This is not the same value as Single.Epsilon
#endregion Private Fields
#region Public Methods
public static Sym3x3 ComputeWeightedCovariance(int n, Vec3[] points, float[] weights) {
//
// compute the centroid
//
float total = 0.0f;
Vec3 centroid = new Vec3(0.0f);
for(int i = 0; i < n; ++i) {
total += weights[i];
centroid += weights[i] * points[i];
}
centroid /= total;
//
// accumulate the covariance matrix
//
Sym3x3 covariance = new Sym3x3(0.0f);
for(int i = 0; i < n; ++i) {
Vec3 a = points[i] - centroid;
Vec3 b = weights[i] * a;
covariance[0] += a.X * b.X;
covariance[1] += a.X * b.Y;
covariance[2] += a.X * b.Z;
covariance[3] += a.Y * b.Y;
covariance[4] += a.Y * b.Z;
covariance[5] += a.Z * b.Z;
}
//
// return it
//
return covariance;
}
public static Vec3 ComputePrincipleComponent(Sym3x3 matrix)
{
//
// compute the cubic coefficients
//
float c0 = matrix[0] * matrix[3] * matrix[5] + 2.0f * matrix[1] * matrix[2] * matrix[4] - matrix[0] * matrix[4] * matrix[4] - matrix[3] * matrix[2] * matrix[2] - matrix[5] * matrix[1] * matrix[1];
float c1 = matrix[0] * matrix[3] + matrix[0] * matrix[5] + matrix[3] * matrix[5] - matrix[1] * matrix[1] - matrix[2] * matrix[2] - matrix[4] * matrix[4];
float c2 = matrix[0] + matrix[3] + matrix[5];
//
// compute the quadratic coefficients
//
float a = c1 - 1.0f / 3.0f * c2 * c2;
float b = -2.0f / 27.0f * c2 * c2 * c2 + 1.0f / 3.0f * c1 * c2 - c0;
//
// compute the root count check
//
float Q = 0.25f * b * b + 1.0f / 27.0f * a * a * a;
//
// test the multiplicity
//
if (FLT_EPSILON < Q)
{
return(new Vec3(1.0f)); // only one root, which implies we have a multiple of the identity
}
if (Q < -FLT_EPSILON)
{
//
// three distinct roots
//
float theta = (float)Math.Atan2(Math.Sqrt(-Q), -0.5f * b);
float rho = (float)Math.Sqrt(0.25f * b * b - Q);
float rt = (float)Math.Pow(rho, 1.0f / 3.0f);
float ct = (float)Math.Cos(theta / 3.0f);
float st = (float)Math.Sin(theta / 3.0f);
float l1 = 1.0f / 3.0f * c2 + 2.0f * rt * ct;
float l2 = 1.0f / 3.0f * c2 - rt * (ct + (float)Math.Sqrt(3.0f) * st);
float l3 = 1.0f / 3.0f * c2 - rt * (ct - (float)Math.Sqrt(3.0f) * st);
//
// pick the larger
//
if (Math.Abs(l2) > Math.Abs(l1))
{
l1 = l2;
}
if (Math.Abs(l3) > Math.Abs(l1))
{
l1 = l3;
}
//
// get the eigenvector
//
return(GetMultiplicity1Evector(matrix, l1));
}
else
{
//
// two roots
//
float rt;
if (b < 0.0f)
{
rt = -(float)Math.Pow(-0.5f * b, 1.0f / 3.0f);
}
else
{
rt = (float)Math.Pow(0.5f * b, 1.0f / 3.0f);
}
float l1 = 1.0f / 3.0f * c2 + rt; // repeated
float l2 = 1.0f / 3.0f * c2 - 2.0f * rt;
//
// get the eigenvector
//
return(Math.Abs(l1) > Math.Abs(l2) ? GetMultiplicity2Evector(matrix, l1) : GetMultiplicity1Evector(matrix, l2));
}
}
public ClusterFit(ColourSet Colours, SquishFlags Flags) : base(Colours, Flags)
{
//
// initialise the metric
//
bool isPerceptual = (flags & SquishFlags.ColourMetricPerceptual) != 0;
metric = isPerceptual ? new Vec3(0.2126f, 0.7152f, 0.0722f) : new Vec3(1.0f);
//
// initialise the best error
//
bestError = Single.MaxValue;
//
// cache some values
//
int count = colours.Count;
Vec3[] values = colours.Points;
float[] weights = colours.Weights;
//
// get the covariance matrix
//
Sym3x3 covariance = Maths.ComputeWeightedCovariance(count, values, weights);
//
// compute the principle component
//
Vec3 principle = Maths.ComputePrincipleComponent(covariance);
//
// get the min and max range as the codebook endpoints
//
Vec3 startTemp = new Vec3(0.0f);
Vec3 endTemp = new Vec3(0.0f);
if (count > 0)
{
//
// compute the range
//
startTemp = endTemp = values[0];
float min = Vec3.Dot(values[0], principle);
float max = min;
for (int i = 1; i < count; ++i)
{
float val = Vec3.Dot(values[i], principle);
if (val < min)
{
startTemp = values[i];
min = val;
}
else
{
if (val > max)
{
endTemp = values[i];
max = val;
}
}
}
}
//
// clamp the output to [0, 1]
//
Vec3 one = new Vec3(1.0f);
Vec3 zero = new Vec3(0.0f);
startTemp = Vec3.Min(one, Vec3.Max(zero, startTemp));
endTemp = Vec3.Min(one, Vec3.Max(zero, endTemp));
//
// clamp to the grid and save
//
Vec3 grid = new Vec3(31.0f, 63.0f, 31.0f);
Vec3 gridrcp = new Vec3(1.0f / 31.0f, 1.0f / 63.0f, 1.0f / 31.0f);
Vec3 half = new Vec3(0.5f);
start = Vec3.Truncate(grid * startTemp + half) * gridrcp;
end = Vec3.Truncate(grid * endTemp + half) * gridrcp;
}
public static Vec3 ComputePrincipleComponent(Sym3x3 matrix) {
//
// compute the cubic coefficients
//
float c0 = matrix[0] * matrix[3] * matrix[5] + 2.0f * matrix[1] * matrix[2] * matrix[4] - matrix[0] * matrix[4] * matrix[4] - matrix[3] * matrix[2] * matrix[2] - matrix[5] * matrix[1] * matrix[1];
float c1 = matrix[0] * matrix[3] + matrix[0] * matrix[5] + matrix[3] * matrix[5] - matrix[1] * matrix[1] - matrix[2] * matrix[2] - matrix[4] * matrix[4];
float c2 = matrix[0] + matrix[3] + matrix[5];
//
// compute the quadratic coefficients
//
float a = c1 - 1.0f / 3.0f * c2 * c2;
float b = -2.0f / 27.0f * c2 * c2 * c2 + 1.0f / 3.0f * c1 * c2 - c0;
//
// compute the root count check
//
float Q = 0.25f * b * b + 1.0f / 27.0f * a * a * a;
//
// test the multiplicity
//
if (FLT_EPSILON < Q) return new Vec3(1.0f); // only one root, which implies we have a multiple of the identity
if (Q < -FLT_EPSILON) {
//
// three distinct roots
//
float theta = (float)Math.Atan2(Math.Sqrt(-Q), -0.5f * b);
float rho = (float)Math.Sqrt(0.25f * b * b - Q);
float rt = (float)Math.Pow(rho, 1.0f / 3.0f);
float ct = (float)Math.Cos(theta / 3.0f);
float st = (float)Math.Sin(theta / 3.0f);
float l1 = 1.0f / 3.0f * c2 + 2.0f * rt * ct;
float l2 = 1.0f / 3.0f * c2 - rt * (ct + (float)Math.Sqrt(3.0f) * st);
float l3 = 1.0f / 3.0f * c2 - rt * (ct - (float)Math.Sqrt(3.0f) * st);
//
// pick the larger
//
if (Math.Abs(l2) > Math.Abs(l1)) l1 = l2;
if (Math.Abs(l3) > Math.Abs(l1)) l1 = l3;
//
// get the eigenvector
//
return GetMultiplicity1Evector(matrix, l1);
}
else {
//
// two roots
//
float rt;
if (b < 0.0f) rt = -(float)Math.Pow(-0.5f * b, 1.0f / 3.0f);
else rt = (float)Math.Pow(0.5f * b, 1.0f / 3.0f);
float l1 = 1.0f / 3.0f *c2 + rt; // repeated
float l2 = 1.0f / 3.0f *c2 - 2.0f * rt;
//
// get the eigenvector
//
return Math.Abs(l1) > Math.Abs(l2) ? GetMultiplicity2Evector(matrix, l1) : GetMultiplicity1Evector(matrix, l2);
}
}
private static Vec3 GetMultiplicity2Evector(Sym3x3 matrix, float evalue) {
//
// compute M
//
Sym3x3 m = new Sym3x3 {
[0] = matrix[0] - evalue,
[1] = matrix[1],
[2] = matrix[2],
[3] = matrix[3] - evalue,
[4] = matrix[4],
[5] = matrix[5] - evalue
};
//
// find the largest component
//
float mc = Math.Abs(m[0]);
int mi = 0;
for(int i = 1; i < 6; ++i) {
float c = Math.Abs(m[i]);
if (c > mc) {
mc = c;
mi = i;
}
}
//
// pick the first eigenvector based on this index
//
switch(mi) {
case 0:
case 1: return new Vec3(-m[1], m[0], 0.0f);
case 2: return new Vec3(m[2], 0.0f, -m[0]);
case 3:
case 4: return new Vec3(0.0f, -m[4], m[3]);
default: return new Vec3(0.0f, -m[5], m[4]);
}
}
private static Vec3 GetMultiplicity1Evector(Sym3x3 matrix, float evalue) {
//
// compute M
//
Sym3x3 m = new Sym3x3 {
[0] = matrix[0] - evalue,
[1] = matrix[1],
[2] = matrix[2],
[3] = matrix[3] - evalue,
[4] = matrix[4],
[5] = matrix[5] - evalue
};
//
// compute U
//
Sym3x3 u = new Sym3x3 {
[0] = m[3] * m[5] - m[4] * m[4],
[1] = m[2] * m[4] - m[1] * m[5],
[2] = m[1] * m[4] - m[2] * m[3],
[3] = m[0] * m[5] - m[2] * m[2],
[4] = m[1] * m[2] - m[4] * m[0],
[5] = m[0] * m[3] - m[1] * m[1]
};
//
// find the largest component
//
float mc = Math.Abs(u[0]);
int mi = 0;
for(int i = 1; i < 6; ++i) {
float c = Math.Abs(u[i]);
if (c > mc) {
mc = c;
mi = i;
}
}
//
// pick the column with this component
//
switch(mi) {
case 0: return new Vec3(u[0], u[1], u[2]);
case 1:
case 3: return new Vec3(u[1], u[3], u[4]);
default: return new Vec3(u[2], u[4], u[5]);
}
}