feat: DH-19379 Added CountWhere operations (#406)

* Added CountWhere operations

* Remove unnessary num2

* fixed some test method names

* Add more filters to CumCountWhere, remove 0 group and some Incs from others

Author: stanbrub

.github/distro/docker-compose.yml

@@ -5,6 +5,7 @@ services:
       - "${DEEPHAVEN_PORT:-10000}:10000"
     volumes:
       - ./data:/data
+      - ./minio:/minio
     environment:
       - "START_OPTS=-Xmx24G -DAuthHandlers=io.deephaven.auth.AnonymousAuthenticationHandler"
 

src/it/java/io/deephaven/benchmark/tests/standard/aggby/CountWhereTest.java

@@ -0,0 +1,36 @@
+/* Copyright (c) 2025-2026 Deephaven Data Labs and Patent Pending */
+package io.deephaven.benchmark.tests.standard.aggby;
+
+import org.junit.jupiter.api.*;
+import io.deephaven.benchmark.tests.standard.StandardTestRunner;
+
+/**
+ * Standard tests for the countWhere table operation. Returns the number of rows for each group.
+ */
+public class CountWhereTest {
+    final StandardTestRunner runner = new StandardTestRunner(this);
+
+    @BeforeEach
+    void setup() {
+        runner.setRowFactor(6);
+        runner.tables("source");
+
+        var setupStr = "from deephaven import agg";
+        runner.addSetupQuery(setupStr);
+    }
+
+    @Test
+    void countWhere1Group() {
+        runner.setScaleFactors(9, 0);
+        var q = "source.agg_by([agg.count_where(col='count', filters=['num1 > 3'])], by=['key1'])";
+        runner.test("CountWhere-AggBy- Range 1 Group 100 Unique Vals", 100, q, "key1", "num1");
+    }
+
+    @Test
+    void countWhere2Group() {
+        runner.setScaleFactors(3, 2);
+        var q = "source.agg_by([agg.count_where(col='count', filters=['num1 % 3 = 0'])], by=['key1', 'key2'])";
+        runner.test("CountWhere-AggBy- Equals 2 Groups 10K Unique Combos", 10100, q, "key1", "key2", "num1");
+    }
+
+}

src/it/java/io/deephaven/benchmark/tests/standard/updateby/CumCountWhereTest.java

@@ -0,0 +1,66 @@
+/* Copyright (c) 2025-2026 Deephaven Data Labs and Patent Pending */
+package io.deephaven.benchmark.tests.standard.updateby;
+
+import org.junit.jupiter.api.*;
+import io.deephaven.benchmark.tests.standard.StandardTestRunner;
+
+/**
+ * Standard tests for the updateBy table operation. Calculates a cumulative countWhere for specified columns and places
+ * the result into a new column for each row.
+ */
+public class CumCountWhereTest {
+    final StandardTestRunner runner = new StandardTestRunner(this);
+
+    void setup(int rowFactor, int staticFactor, int incFactor) {
+        runner.setRowFactor(rowFactor);
+        runner.setScaleFactors(staticFactor, incFactor);
+        runner.tables("timed");
+        runner.addSetupQuery("""
+        from deephaven.updateby import cum_count_where
+        from deephaven.filters import or_
+        """);
+    }
+
+    @Test
+    void cumCountWhereRange0Group1Col() {
+        setup(5, 10, 0);
+        var q = "timed.update_by(ops=cum_count_where(col='X', filters=['num1 > 3']))";
+        runner.test("CumCountWhere- Range No Groups 1 Col", q, "num1");
+    }
+    
+    @Test
+    void cumCountWhereEquals0Group1Col() {
+        setup(5, 8, 0);
+        var q = "timed.update_by(ops=cum_count_where(col='X', filters=['num1 % 3 = 0']))";
+        runner.test("CumCountWhere- Equals No Groups 1 Col", q, "num1");
+    }
+    
+    @Test
+    void cumCountWhereAnd0Group1Col() {
+        setup(5, 10, 0);
+        var q = "timed.update_by(ops=cum_count_where(col='X', filters=['num1 > 3','num1 % 3 = 0']))";
+        runner.test("CumCountWhere- And No Groups 1 Col", q, "num1");
+    }
+    
+    @Test
+    void cumCountWhereOr0Group1Col() {
+        setup(5, 3, 0);
+        var q = "timed.update_by(ops=cum_count_where(col='X', filters=or_(['num1 > 3','num1 % 3 = 0'])))";
+        runner.test("CumCountWhere- Or No Groups 1 Col", q, "num1");
+    }
+
+    @Test
+    void cumCountWhereRange1Group1Col() {
+        setup(3, 10, 0);
+        var q = "timed.update_by(ops=cum_count_where(col='X', filters=['num1 > 3']), by=['key1'])";
+        runner.test("CumCountWhere- Range 1 Group 100 Unique Vals", q, "key1", "num1");
+    }
+
+    @Test
+    void cumCountWhereEquals2Groups1Col() {
+        setup(2, 3, 1);
+        var q = "timed.update_by(ops=cum_count_where(col='X', filters=['num1 % 3 = 0']), by=['key1','key2'])";
+        runner.test("CumCountWhere- Equals 2 Groups 10K Unique Combos", q, "key1", "key2", "num1");
+    }
+
+}

src/it/java/io/deephaven/benchmark/tests/standard/updateby/RollingCountWhereTickTest.java

@@ -0,0 +1,39 @@
+/* Copyright (c) 2025-2026 Deephaven Data Labs and Patent Pending */
+package io.deephaven.benchmark.tests.standard.updateby;
+
+import org.junit.jupiter.api.*;
+import io.deephaven.benchmark.tests.standard.StandardTestRunner;
+
+/**
+ * Standard tests for the updateBy table operation. Defines a tick-based rolling countWhere. The result table contains
+ * an additional column with windowed rolling countWhere.
+ * <p>
+ * Note: This test must contain benchmarks and <code>rev_ticks/fwd_ticks</code> that are comparable to
+ * <code>RollingCountWhereTimeTest</code>
+ */
+public class RollingCountWhereTickTest {
+    final StandardTestRunner runner = new StandardTestRunner(this);
+    final Setup setup = new Setup(runner);
+    final String fifty = """ 
+        from deephaven.updateby import rolling_count_where_tick
+        contains_row = rolling_count_where_tick('X',filters=["num1 > 3"],rev_ticks=20,fwd_ticks=30)
+        """;
+
+
+    @Test
+    void rollingCountWhereTick1Group3Ops() {
+        setup.factors(5, 4, 0);
+        runner.addSetupQuery(fifty);
+        var q = "timed.update_by(ops=[contains_row], by=['key1'])";
+        runner.test("RollingCountWhereTick- 1 Group 100 Unique Vals", q, "key1", "num1");
+    }
+
+    @Test
+    void rollingCountWhereTick2Groups3Ops() {
+        setup.factors(2, 3, 1);
+        runner.addSetupQuery(fifty);
+        var q = "timed.update_by(ops=[contains_row], by=['key1','key2'])";
+        runner.test("RollingCountWhereTick- 2 Groups 10K Unique Combos", q, "key1", "key2", "num1");
+    }
+
+}

src/it/java/io/deephaven/benchmark/tests/standard/updateby/RollingCountWhereTimeTest.java

@@ -0,0 +1,42 @@
+/* Copyright (c) 2025-2026 Deephaven Data Labs and Patent Pending */
+package io.deephaven.benchmark.tests.standard.updateby;
+
+import org.junit.jupiter.api.*;
+import io.deephaven.benchmark.tests.standard.StandardTestRunner;
+
+/**
+ * Standard tests for the updateBy table operation. Defines a time-based rolling countWhere. The result table contains
+ * an additional column with windowed rolling countWhere.
+ * <p>
+ * Note: This test must contain benchmarks and <code>rev_time/fwd_time</code> that are comparable to
+ * <code>RollingCountWhereTickTest</code>
+ */
+public class RollingCountWhereTimeTest {
+    final StandardTestRunner runner = new StandardTestRunner(this);
+    final Setup setup = new Setup(runner);
+    final String fifty1Group = """
+        from deephaven.updateby import rolling_count_where_time
+        contains_row = rolling_count_where_time("timestamp",'X',filters=["num1 > 3"],rev_time="PT2S",fwd_time="PT3S")
+        """;
+    final String fifty2Groups = """
+        from deephaven.updateby import rolling_count_where_time
+        contains_row = rolling_count_where_time("timestamp",'X',filters=["num1 > 3"],rev_time="PT4M",fwd_time="PT5M")
+        """;
+
+    @Test
+    void rollingCountWhereTime1Group3Ops() {
+        setup.factors(4, 2, 0);
+        runner.addSetupQuery(fifty1Group);
+        var q = "timed.update_by(ops=[contains_row], by=['key1'])";
+        runner.test("RollingCountWhereTime- 1 Group 100 Unique Vals", q, "key1", "num1", "timestamp");
+    }
+
+    @Test
+    void rollingCountWhereTime2Groups3Ops() {
+        setup.factors(2, 2, 1);
+        runner.addSetupQuery(fifty2Groups);
+        var q = "timed.update_by(ops=[contains_row], by=['key1','key2'])";
+        runner.test("RollingCountWhereTime- 2 Groups 10K Unique Combos", q, "key1", "key2", "num1", "timestamp");
+    }
+
+}

src/main/resources/io/deephaven/benchmark/run/profile/queries/publish.py

@@ -9,7 +9,7 @@
 with urlopen(root + '/deephaven-benchmark/benchmark_tables.dh.py') as r:
     benchmark_storage_uri_arg = root + '/deephaven-benchmark'
     benchmark_category_arg = 'nightly'  # release | nightly
-    benchnark_max_sets_arg = 45
+    benchmark_max_sets_arg = 45
     benchmark_actor_filter_arg = 'deephaven'
     benchmark_set_filter_arg = '[0-9]{4}([-][0-9]{2}){2}'  # yyyy-mm-dd
     exec(r.read().decode(), globals(), locals())