kscience/kmath
9
0
Fork 1
Code Issues Pull Requests 6 Packages Projects Releases 16 Wiki Activity

Golub-Kahan SVD algorithm for KMP tensors #499

Closed
grinisrit wants to merge 64 commits from dev into dev
pull from: dev
merge into: kscience:dev
Conversation 1 Commits 64 Files Changed 9 -56
1 changed files with 0 additions and 56 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit 7b693270f0 - Show all commits

56
benchmarks/src/jvmMain/kotlin/space/kscience/kmath/benchmarks/SVDBenchmark.kt
View File

@ -29,13 +29,6 @@ class SVDBenchmark {
@Benchmark @Benchmark
fun svdPowerMethodSmall(blackhole: Blackhole) { fun svdPowerMethodSmall(blackhole: Blackhole) {
val svd = tensorSmall.svdPowerMethod()
val tensorSVD = svd.first
.dot(
diagonalEmbedding(svd.second)
.dot(svd.third.transpose())
)
assertTrue(tensorSVD.eq(tensorSmall, epsilon))
blackhole.consume( blackhole.consume(
tensorSmall.svdPowerMethod() tensorSmall.svdPowerMethod()
) )
@ -43,13 +36,6 @@ class SVDBenchmark {
@Benchmark @Benchmark
fun svdPowerMethodMedium(blackhole: Blackhole) { fun svdPowerMethodMedium(blackhole: Blackhole) {
val svd = tensorMedium.svdPowerMethod()
val tensorSVD = svd.first
.dot(
diagonalEmbedding(svd.second)
.dot(svd.third.transpose())
)
assertTrue(tensorSVD.eq(tensorMedium, epsilon))
blackhole.consume( blackhole.consume(
tensorMedium.svdPowerMethod() tensorMedium.svdPowerMethod()
) )
@ -57,13 +43,6 @@ class SVDBenchmark {
@Benchmark @Benchmark
fun svdPowerMethodLarge(blackhole: Blackhole) { fun svdPowerMethodLarge(blackhole: Blackhole) {
val svd = tensorLarge.svdPowerMethod()
val tensorSVD = svd.first
.dot(
diagonalEmbedding(svd.second)
.dot(svd.third.transpose())
)
assertTrue(tensorSVD.eq(tensorLarge, epsilon))
blackhole.consume( blackhole.consume(
tensorLarge.svdPowerMethod() tensorLarge.svdPowerMethod()
) )
@ -71,13 +50,6 @@ class SVDBenchmark {
@Benchmark @Benchmark
fun svdPowerMethodVeryLarge(blackhole: Blackhole) { fun svdPowerMethodVeryLarge(blackhole: Blackhole) {
val svd = tensorVeryLarge.svdPowerMethod()
val tensorSVD = svd.first
.dot(
diagonalEmbedding(svd.second)
.dot(svd.third.transpose())
)
assertTrue(tensorSVD.eq(tensorVeryLarge, epsilon))
blackhole.consume( blackhole.consume(
tensorVeryLarge.svdPowerMethod() tensorVeryLarge.svdPowerMethod()
) )
@ -85,13 +57,6 @@ class SVDBenchmark {
@Benchmark @Benchmark
fun svdGolubKahanSmall(blackhole: Blackhole) { fun svdGolubKahanSmall(blackhole: Blackhole) {
val svd = tensorSmall.svdGolubKahan()
val tensorSVD = svd.first
.dot(
diagonalEmbedding(svd.second)
.dot(svd.third.transpose())
)
assertTrue(tensorSVD.eq(tensorSmall, epsilon))
blackhole.consume( blackhole.consume(
tensorSmall.svdGolubKahan() tensorSmall.svdGolubKahan()
) )
@ -99,13 +64,6 @@ class SVDBenchmark {
@Benchmark @Benchmark
fun svdGolubKahanMedium(blackhole: Blackhole) { fun svdGolubKahanMedium(blackhole: Blackhole) {
val svd = tensorMedium.svdGolubKahan()
val tensorSVD = svd.first
.dot(
diagonalEmbedding(svd.second)
.dot(svd.third.transpose())
)
assertTrue(tensorSVD.eq(tensorMedium, epsilon))
blackhole.consume( blackhole.consume(
tensorMedium.svdGolubKahan() tensorMedium.svdGolubKahan()
) )
@ -113,13 +71,6 @@ class SVDBenchmark {
@Benchmark @Benchmark
fun svdGolubKahanLarge(blackhole: Blackhole) { fun svdGolubKahanLarge(blackhole: Blackhole) {
val svd = tensorLarge.svdGolubKahan()
val tensorSVD = svd.first
.dot(
diagonalEmbedding(svd.second)
.dot(svd.third.transpose())
)
assertTrue(tensorSVD.eq(tensorLarge, epsilon))
blackhole.consume( blackhole.consume(
tensorLarge.svdGolubKahan() tensorLarge.svdGolubKahan()
) )
@ -127,13 +78,6 @@ class SVDBenchmark {
@Benchmark @Benchmark
fun svdGolubKahanVeryLarge(blackhole: Blackhole) { fun svdGolubKahanVeryLarge(blackhole: Blackhole) {
val svd = tensorVeryLarge.svdGolubKahan()
val tensorSVD = svd.first
.dot(
diagonalEmbedding(svd.second)
.dot(svd.third.transpose())
)
assertTrue(tensorSVD.eq(tensorVeryLarge, epsilon))
blackhole.consume( blackhole.consume(
tensorVeryLarge.svdGolubKahan() tensorVeryLarge.svdGolubKahan()
) )