Skip to content

Add experimental noisyCardinality function #22087

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
pranjalssh merged 1 commit into from
Mar 8, 2024
Merged

Add experimental noisyCardinality function #22087

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 14 additions & 0 deletions presto-main/src/main/java/com/facebook/presto/operator/scalar/HyperLogLogFunctions.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,7 @@
package com.facebook.presto.operator.scalar;

import com.facebook.airlift.stats.cardinality.HyperLogLog;
import com.facebook.airlift.stats.cardinality.PrivateLpcaSketch;
import com.facebook.presto.common.block.Block;
import com.facebook.presto.common.type.StandardTypes;
import com.facebook.presto.spi.function.Description;
Expand All @@ -24,6 +25,9 @@

import static com.facebook.presto.operator.aggregation.ApproximateSetAggregation.DEFAULT_STANDARD_ERROR;
import static com.facebook.presto.operator.aggregation.HyperLogLogUtils.standardErrorToBuckets;
import static com.facebook.presto.spi.StandardErrorCode.INVALID_FUNCTION_ARGUMENT;
import static com.facebook.presto.spi.function.SqlFunctionVisibility.EXPERIMENTAL;
import static com.facebook.presto.util.Failures.checkCondition;

public final class HyperLogLogFunctions
{
Expand All @@ -37,6 +41,16 @@ public static long cardinality(@SqlType(StandardTypes.HYPER_LOG_LOG) Slice seria
return HyperLogLog.newInstance(serializedHll).cardinality();
}

@ScalarFunction(visibility = EXPERIMENTAL)
@Description("compute the noisy cardinality of a HyperLogLog instance")
@SqlType(StandardTypes.BIGINT)
public static long noisyCardinality(@SqlType(StandardTypes.HYPER_LOG_LOG) Slice serializedHll, @SqlType(StandardTypes.DOUBLE) double epsilon)
{
checkCondition(epsilon > 0, INVALID_FUNCTION_ARGUMENT, "Epsilon must be greater than 0");
PrivateLpcaSketch privacySketch = new PrivateLpcaSketch(HyperLogLog.newInstance(serializedHll), 0.1 * epsilon, 0.9 * epsilon);
return privacySketch.cardinality();
}

@ScalarFunction
@Description("an empty HyperLogLog instance")
@SqlType(StandardTypes.HYPER_LOG_LOG)
Expand Down
23 changes: 23 additions & 0 deletions presto-main/src/test/java/com/facebook/presto/operator/scalar/TestHyperLogLogFunctions.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -91,6 +91,21 @@ public void testCardinalitySingleColumn()
functionAssertions.assertFunctionWithError(projection, BIGINT, uniqueElements, error);
}

@Test
public void testNoisyCardinalitySingleColumn()
{
int[] uniqueElementsCount = {0, 10000, 100000, 1000000};
for (int uniqueElements : uniqueElementsCount) {
double error = uniqueElements * 0.05;
HyperLogLog hll = HyperLogLog.newInstance(NUMBER_OF_BUCKETS);
for (int i = 0; i < uniqueElements; i++) {
hll.add(i);
}
String projection = getNoisyCardinalityProjection(hll);
functionAssertions.assertFunctionWithError(projection, BIGINT, uniqueElements, error);
}
}

@Test
public void testCardinalityTwoColumns()
{
Expand Down Expand Up @@ -206,4 +221,12 @@ private String getMergeProjection(List<HyperLogLog> list)

return projection;
}

private String getNoisyCardinalityProjection(HyperLogLog hll)
{
Slice serializedHll = hll.serialize();
byte[] hllBytes = serializedHll.getBytes();
String encodedHll = BaseEncoding.base16().lowerCase().encode(hllBytes);
return String.format("noisy_cardinality(CAST(X'%s' AS HyperLogLog), infinity())", encodedHll);
}
}