liuyu/pgvector
1
0
Fork 0
mirror of https://github.com/pgvector/pgvector.git synced 2026-07-11 15:16:54 +08:00
Code Issues Packages Projects Releases Wiki Activity

Moved code to separate function [skip ci]

Browse Source
This commit is contained in:
Andrew Kane
2024-04-11 22:23:16 -07:00
parent b9d5269547
commit 3621a84ef8
1 changed files with 78 additions and 65 deletions
Download Patch File Download Diff File Expand all files Collapse all files

143
src/ivfkmeans.c
View File

@@ -218,6 +218,83 @@ ShowMemoryUsage(MemoryContext context, Size estimatedSize)
} }
#endif #endif
/*
* Compute new centers
*/
static void
ComputeNewCenters(VectorArray samples, VectorArray aggCenters, int *centerCounts, int *closestCenters, IvfflatType type)
{
int dimensions = aggCenters->dim;
int numCenters = aggCenters->maxlen;
int numSamples = samples->length;
for (int j = 0; j < numCenters; j++)
{
Vector *vec = (Vector *) VectorArrayGet(aggCenters, j);
for (int k = 0; k < dimensions; k++)
vec->x[k] = 0.0;
centerCounts[j] = 0;
}
/* Increment sum of closest center */
if (type == IVFFLAT_TYPE_VECTOR)
{
for (int j = 0; j < numSamples; j++)
{
Vector *aggCenter = (Vector *) VectorArrayGet(aggCenters, closestCenters[j]);
Vector *vec = (Vector *) VectorArrayGet(samples, j);
for (int k = 0; k < dimensions; k++)
aggCenter->x[k] += vec->x[k];
}
}
else if (type == IVFFLAT_TYPE_HALFVEC)
{
for (int j = 0; j < numSamples; j++)
{
Vector *aggCenter = (Vector *) VectorArrayGet(aggCenters, closestCenters[j]);
HalfVector *vec = (HalfVector *) VectorArrayGet(samples, j);
for (int k = 0; k < dimensions; k++)
aggCenter->x[k] += HalfToFloat4(vec->x[k]);
}
}
else
elog(ERROR, "Unsupported type");
/* Increment count of closest center */
for (int j = 0; j < numSamples; j++)
centerCounts[closestCenters[j]] += 1;
/* Average centers */
for (int j = 0; j < numCenters; j++)
{
Vector *vec = (Vector *) VectorArrayGet(aggCenters, j);
if (centerCounts[j] > 0)
{
/* Double avoids overflow, but requires more memory */
/* TODO Update bounds */
for (int k = 0; k < dimensions; k++)
{
if (isinf(vec->x[k]))
vec->x[k] = vec->x[k] > 0 ? FLT_MAX : -FLT_MAX;
}
for (int k = 0; k < dimensions; k++)
vec->x[k] /= centerCounts[j];
}
else
{
/* TODO Handle empty centers properly */
for (int k = 0; k < dimensions; k++)
vec->x[k] = RandomDouble();
}
}
}
/* /*
* Use Elkan for performance. This requires distance function to satisfy triangle inequality. * Use Elkan for performance. This requires distance function to satisfy triangle inequality.
* *
@@ -463,71 +540,7 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers, IvfflatTyp
} }
/* Step 4: For each center c, let m(c) be mean of all points assigned */ /* Step 4: For each center c, let m(c) be mean of all points assigned */
for (int j = 0; j < numCenters; j++) ComputeNewCenters(samples, aggCenters, centerCounts, closestCenters, type);
{
Vector *vec = (Vector *) VectorArrayGet(aggCenters, j);
for (int k = 0; k < dimensions; k++)
vec->x[k] = 0.0;
centerCounts[j] = 0;
}
/* Increment sum of closest center */
if (type == IVFFLAT_TYPE_VECTOR)
{
for (int j = 0; j < numSamples; j++)
{
Vector *aggCenter = (Vector *) VectorArrayGet(aggCenters, closestCenters[j]);
Vector *vec = (Vector *) VectorArrayGet(samples, j);
for (int k = 0; k < dimensions; k++)
aggCenter->x[k] += vec->x[k];
}
}
else if (type == IVFFLAT_TYPE_HALFVEC)
{
for (int j = 0; j < numSamples; j++)
{
Vector *aggCenter = (Vector *) VectorArrayGet(aggCenters, closestCenters[j]);
HalfVector *vec = (HalfVector *) VectorArrayGet(samples, j);
for (int k = 0; k < dimensions; k++)
aggCenter->x[k] += HalfToFloat4(vec->x[k]);
}
}
else
elog(ERROR, "Unsupported type");
/* Increment count of closest center */
for (int j = 0; j < numSamples; j++)
centerCounts[closestCenters[j]] += 1;
/* Average centers */
for (int j = 0; j < numCenters; j++)
{
Vector *vec = (Vector *) VectorArrayGet(aggCenters, j);
if (centerCounts[j] > 0)
{
/* Double avoids overflow, but requires more memory */
/* TODO Update bounds */
for (int k = 0; k < dimensions; k++)
{
if (isinf(vec->x[k]))
vec->x[k] = vec->x[k] > 0 ? FLT_MAX : -FLT_MAX;
}
for (int k = 0; k < dimensions; k++)
vec->x[k] /= centerCounts[j];
}
else
{
/* TODO Handle empty centers properly */
for (int k = 0; k < dimensions; k++)
vec->x[k] = RandomDouble();
}
}
/* Set new centers if different from agg centers */ /* Set new centers if different from agg centers */
if (type == IVFFLAT_TYPE_HALFVEC) if (type == IVFFLAT_TYPE_HALFVEC)