[SEDONA-658] Add ST_Simplify (#1606)

* feat: Add ST_Simplify

* fix: snowflake tests

* docs: fix formatting

common/src/main/java/org/apache/sedona/common/Functions.java

@@ -60,6 +60,7 @@
 import org.locationtech.jts.operation.valid.TopologyValidationError;
 import org.locationtech.jts.precision.GeometryPrecisionReducer;
 import org.locationtech.jts.precision.MinimumClearance;
+import org.locationtech.jts.simplify.DouglasPeuckerSimplifier;
 import org.locationtech.jts.simplify.PolygonHullSimplifier;
 import org.locationtech.jts.simplify.TopologyPreservingSimplifier;
 import org.locationtech.jts.simplify.VWSimplifier;
@@ -1543,6 +1544,10 @@ public static Geometry[] h3ToGeom(long[] cells) {
     return polygons.toArray(new Polygon[0]);
   }
 
+  public static Geometry simplify(Geometry geom, double distanceTolerance) {
+    return DouglasPeuckerSimplifier.simplify(geom, distanceTolerance);
+  }
+
   // create static function named simplifyPreserveTopology
   public static Geometry simplifyPreserveTopology(Geometry geometry, double distanceTolerance) {
     return TopologyPreservingSimplifier.simplify(geometry, distanceTolerance);

common/src/test/java/org/apache/sedona/common/FunctionsTest.java

@@ -1790,6 +1790,29 @@ public void removeRepeatedPointsGeometryCollection() throws ParseException {
     assertEquals(6000, actualSRID);
   }
 
+  @Test
+  public void simplify() throws ParseException {
+    Geometry geom = Constructors.geomFromWKT("POINT (1 2)", 1111);
+    geom = Functions.buffer(geom, 10, false, "quad_segs=12");
+    int actualPoints = Functions.nPoints(Functions.simplify(geom, 0.1));
+    int expectedPoints = 33;
+    assertEquals(expectedPoints, actualPoints);
+
+    actualPoints = Functions.nPoints(Functions.simplify(geom, 0.5));
+    expectedPoints = 17;
+    assertEquals(expectedPoints, actualPoints);
+
+    actualPoints = Functions.nPoints(Functions.simplify(geom, 1));
+    expectedPoints = 9;
+    assertEquals(expectedPoints, actualPoints);
+
+    Geometry actual = Functions.simplify(geom, 10);
+    actualPoints = Functions.nPoints(actual);
+    expectedPoints = 4;
+    assertEquals(expectedPoints, actualPoints);
+    assertEquals(1111, actual.getSRID());
+  }
+
   @Test
   public void simplifyVW() throws ParseException {
     Geometry geom = Constructors.geomFromEWKT("LINESTRING(5 2, 3 8, 6 20, 7 25, 10 10)");

docs/api/flink/Function.md

@@ -3513,6 +3513,29 @@ Output:
 3021
 ```
 
+## ST_Simplify
+
+Introduction: This function simplifies the input geometry by applying the Douglas-Peucker algorithm.
+
+!!!Note
+    The simplification may not preserve topology, potentially producing invalid geometries. Use [ST_SimplifyPreserveTopology](#st_simplifypreservetopology) to retain valid topology after simplification.
+
+Format: `ST_Simplify(geom: Geometry, tolerance: Double)`
+
+Since: `v1.7.0`
+
+SQL Example:
+
+```sql
+SELECT ST_Simplify(ST_Buffer(ST_GeomFromWKT('POINT (0 2)'), 10), 1)
+```
+
+Output:
+
+```
+POLYGON ((10 2, 7.0710678118654755 -5.071067811865475, 0.0000000000000006 -8, -7.071067811865475 -5.0710678118654755, -10 1.9999999999999987, -7.071067811865477 9.071067811865476, -0.0000000000000018 12, 7.071067811865474 9.071067811865477, 10 2))
+```
+
 ## ST_SimplifyPolygonHull
 
 Introduction: This function computes a topology-preserving simplified hull, either outer or inner, for a polygonal geometry input. An outer hull fully encloses the original geometry, while an inner hull lies entirely within. The result maintains the same structure as the input, including handling of MultiPolygons and holes, represented as a polygonal geometry formed from a subset of vertices.

docs/api/snowflake/vector-data/Function.md

@@ -2698,6 +2698,27 @@ Output:
 LINESTRING(177 10, 179 10, 181 10, 183 10)
 ```
 
+## ST_Simplify
+
+Introduction: This function simplifies the input geometry by applying the Douglas-Peucker algorithm.
+
+!!!Note
+    The simplification may not preserve topology, potentially producing invalid geometries. Use [ST_SimplifyPreserveTopology](#st_simplifypreservetopology) to retain valid topology after simplification.
+
+Format: `ST_Simplify(geom: Geometry, tolerance: Double)`
+
+SQL Example:
+
+```sql
+SELECT ST_Simplify(ST_Buffer(ST_GeomFromWKT('POINT (0 2)'), 10), 1)
+```
+
+Output:
+
+```
+POLYGON ((10 2, 7.0710678118654755 -5.071067811865475, 0.0000000000000006 -8, -7.071067811865475 -5.0710678118654755, -10 1.9999999999999987, -7.071067811865477 9.071067811865476, -0.0000000000000018 12, 7.071067811865474 9.071067811865477, 10 2))
+```
+
 ## ST_SimplifyPolygonHull
 
 Introduction: This function computes a topology-preserving simplified hull, either outer or inner, for a polygonal geometry input. An outer hull fully encloses the original geometry, while an inner hull lies entirely within. The result maintains the same structure as the input, including handling of MultiPolygons and holes, represented as a polygonal geometry formed from a subset of vertices.

docs/api/sql/Function.md

@@ -3573,6 +3573,29 @@ Output:
 LINESTRING(177 10, 179 10, 181 10, 183 10)
 ```
 
+## ST_Simplify
+
+Introduction: This function simplifies the input geometry by applying the Douglas-Peucker algorithm.
+
+!!!Note
+    The simplification may not preserve topology, potentially producing invalid geometries. Use [ST_SimplifyPreserveTopology](#st_simplifypreservetopology) to retain valid topology after simplification.
+
+Format: `ST_Simplify(geom: Geometry, tolerance: Double)`
+
+Since: `v1.7.0`
+
+SQL Example:
+
+```sql
+SELECT ST_Simplify(ST_Buffer(ST_GeomFromWKT('POINT (0 2)'), 10), 1)
+```
+
+Output:
+
+```
+POLYGON ((10 2, 7.0710678118654755 -5.071067811865475, 0.0000000000000006 -8, -7.071067811865475 -5.0710678118654755, -10 1.9999999999999987, -7.071067811865477 9.071067811865476, -0.0000000000000018 12, 7.071067811865474 9.071067811865477, 10 2))
+```
+
 ## ST_SimplifyPolygonHull
 
 Introduction: This function computes a topology-preserving simplified hull, either outer or inner, for a polygonal geometry input. An outer hull fully encloses the original geometry, while an inner hull lies entirely within. The result maintains the same structure as the input, including handling of MultiPolygons and holes, represented as a polygonal geometry formed from a subset of vertices.

flink/src/main/java/org/apache/sedona/flink/Catalog.java

@@ -172,6 +172,7 @@ public static UserDefinedFunction[] getFuncs() {
       new Functions.ST_Multi(),
       new Functions.ST_StartPoint(),
       new Functions.ST_ShiftLongitude(),
+      new Functions.ST_Simplify(),
       new Functions.ST_SimplifyPreserveTopology(),
       new Functions.ST_SimplifyVW(),
       new Functions.ST_SimplifyPolygonHull(),

flink/src/main/java/org/apache/sedona/flink/expressions/Functions.java

@@ -1394,6 +1394,16 @@ public Geometry[] eval(@DataTypeHint(value = "ARRAY<BIGINT>") Long[] cells) {
     }
   }
 
+  public static class ST_Simplify extends ScalarFunction {
+    @DataTypeHint(value = "RAW", bridgedTo = org.locationtech.jts.geom.Geometry.class)
+    public Geometry eval(
+        @DataTypeHint(value = "RAW", bridgedTo = org.locationtech.jts.geom.Geometry.class) Object o,
+        @DataTypeHint("Double") Double distanceTolerance) {
+      Geometry geom = (Geometry) o;
+      return org.apache.sedona.common.Functions.simplify(geom, distanceTolerance);
+    }
+  }
+
   public static class ST_SimplifyPreserveTopology extends ScalarFunction {
     @DataTypeHint(value = "RAW", bridgedTo = org.locationtech.jts.geom.Geometry.class)
     public Geometry eval(

flink/src/test/java/org/apache/sedona/flink/FunctionTest.java

@@ -1739,6 +1739,17 @@ public void testStartPoint() {
     assertEquals("POINT (0 0)", result.toString());
   }
 
+  @Test
+  public void testSimplify() {
+    Table table = tableEnv.sqlQuery("SELECT ST_Buffer(ST_GeomFromWKT('POINT (0 2)'), 10) AS geom");
+    Geometry actualGeometry =
+        (Geometry)
+            first(table.select(call(Functions.ST_Simplify.class.getSimpleName(), $("geom"), 1)))
+                .getField(0);
+    int actualPoints = actualGeometry.getNumPoints();
+    assertEquals(9, actualPoints);
+  }
+
   @Test
   public void testSimplifyPreserveTopology() {
     Table table =

python/sedona/sql/st_functions.py

@@ -1666,6 +1666,22 @@ def ST_SubDivideExplode(
     return _call_st_function("ST_SubDivideExplode", (geometry, max_vertices))
 
 
+@validate_argument_types
+def ST_Simplify(
+    geometry: ColumnOrName, distance_tolerance: ColumnOrNameOrNumber
+) -> Column:
+    """Simplify a geometry using Douglas-Peucker algorithm within a specified tolerance while preserving topological relationships.
+
+    :param geometry: Geometry column to simplify.
+    :type geometry: ColumnOrName
+    :param distance_tolerance: Tolerance for merging points together to simplify the geometry as either a number or numeric column.
+    :type distance_tolerance: ColumnOrNameOrNumber
+    :return: Simplified geometry as a geometry column.
+    :rtype: Column
+    """
+    return _call_st_function("ST_Simplify", (geometry, distance_tolerance))
+
+
 @validate_argument_types
 def ST_SimplifyPreserveTopology(
     geometry: ColumnOrName, distance_tolerance: ColumnOrNameOrNumber

python/tests/sql/test_dataframe_api.py

@@ -900,6 +900,13 @@
         "",
         "POLYGON ((0 0, 1 0, 1 1, 0 0))",
     ),
+    (
+        stf.ST_Simplify,
+        ("geom", 0.1),
+        "0.9_poly",
+        "",
+        "POLYGON ((0 0, 1 0, 1 1, 0 0))",
+    ),
     (
         stf.ST_SimplifyPreserveTopology,
         ("geom", 0.2),
@@ -1330,6 +1337,8 @@
     (stf.ST_SetSRID, ("", None)),
     (stf.ST_SetSRID, ("", 3021.0)),
     (stf.ST_ShiftLongitude, (None,)),
+    (stf.ST_Simplify, (None, 2)),
+    (stf.ST_Simplify, ("", None)),
     (stf.ST_SimplifyPreserveTopology, (None, 0.2)),
     (stf.ST_SimplifyPreserveTopology, ("", None)),
     (stf.ST_SimplifyVW, (None, 2)),

python/tests/sql/test_function.py

@@ -1442,6 +1442,14 @@ def test_isPolygonCW(self):
         ).take(1)[0][0]
         assert actual
 
+    def test_st_simplify(self):
+        baseDf = self.spark.sql(
+            "SELECT ST_Buffer(ST_GeomFromWKT('POINT (0 2)'), 10) AS geom"
+        )
+        actualPoints = baseDf.selectExpr("ST_NPoints(ST_Simplify(geom, 1))").first()[0]
+        expectedPoints = 9
+        assert expectedPoints == actualPoints
+
     def test_st_simplify_vw(self):
         basedf = self.spark.sql(
             "SELECT ST_GeomFromWKT('LINESTRING(5 2, 3 8, 6 20, 7 25, 10 10)') as geom"

snowflake-tester/src/test/java/org/apache/sedona/snowflake/snowsql/TestFunctions.java

@@ -877,6 +877,16 @@ public void test_ST_SetSRID() {
         "SRID=4326;POINT (1 2)");
   }
 
+  @Test
+  public void test_ST_Simplify() {
+    registerUDF("ST_Simplify", byte[].class, double.class);
+    registerUDF("ST_Buffer", byte[].class, double.class);
+    registerUDF("ST_NPoints", byte[].class);
+    verifySqlSingleRes(
+        "SELECT sedona.ST_NPoints(sedona.ST_Simplify(sedona.ST_Buffer(sedona.ST_GeomFromWKT('POINT (0 2)'), 10), 1))",
+        9);
+  }
+
   @Test
   public void test_ST_SimplifyPreserveTopology() {
     registerUDF("ST_SimplifyPreserveTopology", byte[].class, double.class);

snowflake-tester/src/test/java/org/apache/sedona/snowflake/snowsql/TestFunctionsV2.java

@@ -838,6 +838,16 @@ public void test_ST_SetSRID() {
         "SRID=0;POINT(1 2)");
   }
 
+  @Test
+  public void test_ST_Simplify() {
+    registerUDFV2("ST_Simplify", String.class, double.class);
+    registerUDFV2("ST_Buffer", String.class, double.class);
+    registerUDFV2("ST_NPoints", String.class);
+    verifySqlSingleRes(
+        "SELECT sedona.ST_NPoints(sedona.ST_Simplify(sedona.ST_Buffer(ST_GeomFromWKT('POINT (0 2)'), 10), 1))",
+        9);
+  }
+
   @Test
   public void test_ST_SimplifyPreserveTopology() {
     registerUDFV2("ST_SimplifyPreserveTopology", String.class, double.class);

snowflake/src/main/java/org/apache/sedona/snowflake/snowsql/UDFs.java

@@ -967,6 +967,12 @@ public static byte[] ST_SetSRID(byte[] geometry, int srid) {
     return GeometrySerde.serialize(Functions.setSRID(GeometrySerde.deserialize(geometry), srid));
   }
 
+  @UDFAnnotations.ParamMeta(argNames = {"geometry", "distanceTolerance"})
+  public static byte[] ST_Simplify(byte[] geometry, double distanceTolerance) {
+    return GeometrySerde.serialize(
+        Functions.simplify(GeometrySerde.deserialize(geometry), distanceTolerance));
+  }
+
   @UDFAnnotations.ParamMeta(argNames = {"geometry", "distanceTolerance"})
   public static byte[] ST_SimplifyPreserveTopology(byte[] geometry, double distanceTolerance) {
     return GeometrySerde.serialize(

snowflake/src/main/java/org/apache/sedona/snowflake/snowsql/UDFsV2.java

@@ -1090,6 +1090,15 @@ public static String ST_SetSRID(String geometry, int srid) {
     return GeometrySerde.serGeoJson(Functions.setSRID(GeometrySerde.deserGeoJson(geometry), srid));
   }
 
+  @UDFAnnotations.ParamMeta(
+      argNames = {"geometry", "distanceTolerance"},
+      argTypes = {"Geometry", "double"},
+      returnTypes = "Geometry")
+  public static String ST_Simplify(String geometry, double distanceTolerance) {
+    return GeometrySerde.serGeoJson(
+        Functions.simplify(GeometrySerde.deserGeoJson(geometry), distanceTolerance));
+  }
+
   @UDFAnnotations.ParamMeta(
       argNames = {"geometry", "distanceTolerance"},
       argTypes = {"Geometry", "double"},

spark/common/src/main/scala/org/apache/sedona/sql/UDF/Catalog.scala

@@ -116,6 +116,7 @@ object Catalog {
     function[ST_AsHEXEWKB](),
     function[ST_AsGML](),
     function[ST_AsKML](),
+    function[ST_Simplify](),
     function[ST_SimplifyVW](),
     function[ST_SimplifyPolygonHull](),
     function[ST_SRID](),

spark/common/src/main/scala/org/apache/spark/sql/sedona_sql/expressions/Functions.scala

@@ -442,6 +442,14 @@ case class ST_ReducePrecision(inputExpressions: Seq[Expression])
   }
 }
 
+case class ST_Simplify(inputExpressions: Seq[Expression])
+    extends InferredExpression(Functions.simplify _) {
+
+  protected def withNewChildrenInternal(newChildren: IndexedSeq[Expression]) = {
+    copy(inputExpressions = newChildren)
+  }
+}
+
 case class ST_SimplifyVW(inputExpressions: Seq[Expression])
     extends InferredExpression(Functions.simplifyVW _) {
 

spark/common/src/main/scala/org/apache/spark/sql/sedona_sql/expressions/st_functions.scala

@@ -596,6 +596,11 @@ object st_functions extends DataFrameAPI {
   def ST_Transform(geometry: Column, targetCRS: Column): Column =
     wrapExpression[ST_Transform](geometry, targetCRS)
 
+  def ST_Simplify(geometry: Column, distanceTolerance: Column): Column =
+    wrapExpression[ST_Simplify](geometry, distanceTolerance)
+  def ST_Simplify(geometry: String, distanceTolerance: Double): Column =
+    wrapExpression[ST_Simplify](geometry, distanceTolerance)
+
   def ST_SimplifyVW(geometry: Column, distanceTolerance: Column): Column =
     wrapExpression[ST_SimplifyVW](geometry, distanceTolerance)
   def ST_SimplifyVW(geometry: String, distanceTolerance: Double): Column =

spark/common/src/test/scala/org/apache/sedona/sql/PreserveSRIDSuite.scala

@@ -50,6 +50,7 @@ class PreserveSRIDSuite extends TestBaseScala with TableDrivenPropertyChecks {
       ("ST_Intersection(geom1, ST_Point(0, 1))", 1000),
       ("ST_MakeValid(geom1)", 1000),
       ("ST_ReducePrecision(geom1, 6)", 1000),
+      ("ST_Simplify(geom1, 0.1)", 1000),
       ("ST_SimplifyVW(geom1, 0.1)", 1000),
       ("ST_SimplifyPolygonHull(geom1, 0.5)", 1000),
       ("ST_SetSRID(geom1, 2000)", 2000),

spark/common/src/test/scala/org/apache/sedona/sql/dataFrameAPITestScala.scala

@@ -900,6 +900,13 @@ class dataFrameAPITestScala extends TestBaseScala {
       assert(actualResult == expectedResult)
     }
 
+    it("Passed ST_Simplify") {
+      val baseDf = sparkSession.sql("SELECT ST_Buffer(ST_GeomFromWKT('POINT (0 2)'), 10) AS geom")
+      val actualPoints = baseDf.select(ST_NPoints(ST_Simplify("geom", 1))).first().get(0)
+      val expectedPoints = 9
+      assertEquals(expectedPoints, actualPoints)
+    }
+
     it("Passed ST_SimplifyVW") {
       val baseDf = sparkSession.sql(
         "SELECT ST_GeomFromWKT('LINESTRING(5 2, 3 8, 6 20, 7 25, 10 10)') AS geom")

spark/common/src/test/scala/org/apache/sedona/sql/functionTestScala.scala

@@ -884,6 +884,13 @@ class functionTestScala
       assert(Hex.encodeHexString(df.first().get(0).asInstanceOf[Array[Byte]]) == s)
     }
 
+    it("Passed ST_Simplify") {
+      val baseDf = sparkSession.sql("SELECT ST_Buffer(ST_GeomFromWKT('POINT (0 2)'), 10) AS geom")
+      val actualPoints = baseDf.selectExpr("ST_NPoints(ST_Simplify(geom, 1))").first().get(0)
+      val expectedPoints = 9
+      assertEquals(expectedPoints, actualPoints)
+    }
+
     it("Passed ST_SimplifyVW") {
       val baseDf = sparkSession.sql(
         "SELECT ST_GeomFromWKT('LINESTRING(5 2, 3 8, 6 20, 7 25, 10 10)') AS geom")