This is an automated email from the ASF dual-hosted git repository.
placave pushed a commit to branch cpc-sketch
in repository https://gitbox.apache.org/repos/asf/datasketches-go.git
The following commit(s) were added to refs/heads/cpc-sketch by this push:
new 2ee945b expose format, flavor and start update test
2ee945b is described below
commit 2ee945b369ead81570071118ae5323df25e5c5ac
Author: Pierre Lacave <[email protected]>
AuthorDate: Sat Jul 20 19:32:06 2024 +0200
expose format, flavor and start update test
---
cpc/cpc_sketch.go | 10 ++++-----
cpc/cpc_sketch_test.go | 58 +++++++++++++++++++++++++++++++-------------------
cpc/utils.go | 46 +++++++++++++++++++++------------------
3 files changed, 66 insertions(+), 48 deletions(-)
diff --git a/cpc/cpc_sketch.go b/cpc/cpc_sketch.go
index 4d1c157..0f631cd 100644
--- a/cpc/cpc_sketch.go
+++ b/cpc/cpc_sketch.go
@@ -199,10 +199,10 @@ func hash(bs []byte, seed uint64) (uint64, uint64) {
return murmur3.SeedSum128(seed, seed, bs)
}
-func (c *CpcSketch) getFormat() cpcFormat {
+func (c *CpcSketch) getFormat() CpcFormat {
ordinal := 0
- f := c.getFlavor()
- if f == flavor_hybrid || f == flavor_sparse {
+ f := c.GetFlavor()
+ if f == CpcFlavorHybrid || f == CpcFlavorSparse {
ordinal = 2
if !c.mergeFlag {
ordinal |= 1
@@ -219,10 +219,10 @@ func (c *CpcSketch) getFormat() cpcFormat {
ordinal |= 1
}
}
- return cpcFormat(ordinal)
+ return CpcFormat(ordinal)
}
-func (c *CpcSketch) getFlavor() cpcFlavor {
+func (c *CpcSketch) GetFlavor() CpcFlavor {
return determineFlavor(c.lgK, c.numCoupons)
}
diff --git a/cpc/cpc_sketch_test.go b/cpc/cpc_sketch_test.go
index fef126e..cb0661f 100644
--- a/cpc/cpc_sketch_test.go
+++ b/cpc/cpc_sketch_test.go
@@ -25,7 +25,7 @@ import (
func TestCPCCheckUpdatesEstimate(t *testing.T) {
sk, err := NewCpcSketch(10, 0)
assert.NoError(t, err)
- assert.Equal(t, sk.getFormat(), format_empty_hip)
+ assert.Equal(t, sk.getFormat(), CpcFormatEmptyHip)
err = sk.UpdateUint64(1)
assert.NoError(t, err)
err = sk.UpdateFloat64(2.0)
@@ -50,33 +50,47 @@ func TestCPCCheckUpdatesEstimate(t *testing.T) {
assert.True(t, lb >= 0)
assert.True(t, lb <= est)
assert.True(t, est <= ub)
+ assert.Equal(t, sk.GetFlavor(), CpcFlavorSparse)
+ assert.Equal(t, sk.getFormat(), CpcFormatSparceHybridHip)
+}
+
+func TestCPCCheckEstimatesWithMerge(t *testing.T) {
+ lgk := 4
+ sk1, err := NewCpcSketch(lgk, CpcDefaultUpdateSeed)
+ assert.NoError(t, err)
+ sk2, err := NewCpcSketch(lgk, CpcDefaultUpdateSeed)
+ assert.NoError(t, err)
+ n := 1 << lgk
+ for i := 0; i < n; i++ {
+ err = sk1.UpdateUint64(uint64(i))
+ assert.NoError(t, err)
+ err = sk2.UpdateUint64(uint64(i + n))
+ assert.NoError(t, err)
+ }
}
/*
@Test
- public void checkUpdatesEstimate() {
- final CpcSketch sk = new CpcSketch(10, 0);
- println(sk.toString(true));
- assertEquals(sk.getFormat(), Format.EMPTY_HIP);
- sk.update(1L);
- sk.update(2.0);
- sk.update("3");
- byte[] bytes = new byte[] { 4, 4 };
- sk.update(bytes);
- sk.update(ByteBuffer.wrap(bytes)); // same as previous
- sk.update(ByteBuffer.wrap(bytes, 0, 1));
- sk.update(new char[] { 5 });
- sk.update(new int[] { 6 });
- sk.update(new long[] { 7 });
- final double est = sk.getEstimate();
- final double lb = sk.getLowerBound(2);
- final double ub = sk.getUpperBound(2);
+ public void checkEstimatesWithMerge() {
+ final int lgK = 4;
+ final CpcSketch sk1 = new CpcSketch(lgK);
+ final CpcSketch sk2 = new CpcSketch(lgK);
+ final int n = 1 << lgK;
+ for (int i = 0; i < n; i++ ) {
+ sk1.update(i);
+ sk2.update(i + n);
+ }
+ final CpcUnion union = new CpcUnion(lgK);
+ union.update(sk1);
+ union.update(sk2);
+ final CpcSketch result = union.getResult();
+ final double est = result.getEstimate();
+ final double lb = result.getLowerBound(2);
+ final double ub = result.getUpperBound(2);
assertTrue(lb >= 0);
assertTrue(lb <= est);
assertTrue(est <= ub);
- assertEquals(sk.getFlavor(), Flavor.SPARSE);
- assertEquals(sk.getFormat(), Format.SPARSE_HYBRID_HIP);
- println(sk.toString());
- println(sk.toString(true));
+ assertTrue(result.validate());
+ println(result.toString(true));
}
*/
diff --git a/cpc/utils.go b/cpc/utils.go
index 6660d60..5897888 100644
--- a/cpc/utils.go
+++ b/cpc/utils.go
@@ -19,26 +19,30 @@ package cpc
import "fmt"
-type cpcFormat int
-type cpcFlavor int
+type CpcFormat int
+type CpcFlavor int
const (
- format_empty_merged cpcFormat = 0
- format_empty_hip cpcFormat = 1
- format_sparse_hybrid_merged cpcFormat = 2
- format_sparce_hybrid_hip cpcFormat = 3
- format_pinned_sliding_merged_nosv cpcFormat = 4
- format_pinned_sliding_hip_nosv cpcFormat = 5
- format_pinned_sliding_merged cpcFormat = 6
- format_pinned_sliding_hip cpcFormat = 7
+ CpcformatEmptyMerged CpcFormat = 0
+ CpcFormatEmptyHip CpcFormat = 1
+ CpcFormatSparseHybridMerged CpcFormat = 2
+ CpcFormatSparceHybridHip CpcFormat = 3
+ CpcFormatPinnedSlidingMergedNosv CpcFormat = 4
+ CpcFormatPinnedSlidingHipNosv CpcFormat = 5
+ CpcFormatPinnedSlidingMerged CpcFormat = 6
+ CpcFormatPinnedSlidingHip CpcFormat = 7
)
const (
- flavor_empty cpcFlavor = 0 // 0 == C < 1
- flavor_sparse cpcFlavor = 1 // 1 <= C < 3K/32
- flavor_hybrid cpcFlavor = 2 // 3K/32 <= C < K/2
- flavor_pinned cpcFlavor = 3 // K/2 <= C < 27K/8 [NB: 27/8 = 3 + 3/8]
- flavor_sliding cpcFlavor = 4 // 27K/8 <= C
+ CpcFlavorEmpty CpcFlavor = 0 // 0 == C < 1
+ CpcFlavorSparse CpcFlavor = 1 // 1 <= C < 3K/32
+ CpcFlavorHybrid CpcFlavor = 2 // 3K/32 <= C < K/2
+ CpcFlavorPinned CpcFlavor = 3 // K/2 <= C < 27K/8 [NB: 27/8 = 3 +
3/8]
+ CpcFlavorSliding CpcFlavor = 4 // 27K/8 <= C
+)
+
+const (
+ CpcDefaultUpdateSeed = 9001
)
func checkLgK(lgK int) error {
@@ -55,23 +59,23 @@ func checkLgSizeInts(lgSizeInts int) error {
return nil
}
-func determineFlavor(lgK int, numCoupons uint64) cpcFlavor {
+func determineFlavor(lgK int, numCoupons uint64) CpcFlavor {
c := numCoupons
k := uint64(1) << lgK
c2 := c << 1
c8 := c << 3
c32 := c << 5
if c == 0 {
- return flavor_empty // 0 == C < 1
+ return CpcFlavorEmpty // 0 == C < 1
}
if c32 < (3 * k) {
- return flavor_sparse // 1 <= C < 3K/32
+ return CpcFlavorSparse // 1 <= C < 3K/32
}
if c2 < k {
- return flavor_hybrid // 3K/32 <= C < K/2
+ return CpcFlavorHybrid // 3K/32 <= C < K/2
}
if c8 < (27 * k) {
- return flavor_pinned // K/2 <= C < 27K/8
+ return CpcFlavorPinned // K/2 <= C < 27K/8
}
- return flavor_sliding // 27K/8 <= C
+ return CpcFlavorSliding // 27K/8 <= C
}
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
