[SEDONA-439] Add RS_Union_Aggr

common/src/main/java/org/apache/sedona/common/raster/PixelRecord.java

@@ -1,3 +1,21 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *   http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ * KIND, either express or implied.  See the License for the
+ * specific language governing permissions and limitations
+ * under the License.
+ */
 package org.apache.sedona.common.raster;
 
 import org.locationtech.jts.geom.Geometry;

docs/api/sql/Raster-aggregate-function.md

@@ -0,0 +1,42 @@
+## RS_Union_Aggr
+
+Introduction: Returns a raster containing bands by specified indexes from all rasters in the provided column. Extracts the first bands from each raster and combines them into the output raster based on the input index values.
+
+!!!Note
+    RS_Union_Aggr can take multiple banded rasters as input but it would only extract the first band to the resulting raster. RS_Union_Aggr expects the following input, if not satisfied then will throw an IllegalArgumentException:
+
+    - Indexes to be in an arithmetic sequence without any gaps.
+    - Indexes to be unique and not repeated.
+    - Rasters should be of the same shape.
+
+Format: `RS_Union_Aggr(A: rasterColumn, B: indexColumn)`
+
+Since: `v1.5.1`
+
+Spark SQL Example:
+
+Contents of `raster_table`.
+
+```
++------------------------------+-----+
+|                        raster|index|
++------------------------------+-----+
+|GridCoverage2D["geotiff_cov...|    1|
+|GridCoverage2D["geotiff_cov...|    2|
+|GridCoverage2D["geotiff_cov...|    3|
+|GridCoverage2D["geotiff_cov...|    4|
+|GridCoverage2D["geotiff_cov...|    5|
++------------------------------+-----+
+```
+
+```
+SELECT RS_Union_Aggr(raster, index) FROM raster_table
+```
+
+Output:
+
+This output raster contains the first band of each raster in the `raster_table` at specified index.
+
+```
+GridCoverage2D["geotiff_coverage", GeneralEnvel...
+```

mkdocs.yml

@@ -64,6 +64,7 @@ nav:
                   - SedonaKepler: api/sql/Visualization_SedonaKepler.md
           - Raster data:
               - Raster loader: api/sql/Raster-loader.md
+              - Raster aggregates: api/sql/Raster-aggregate-function.md
               - Raster writer: api/sql/Raster-writer.md
               - Raster operators: api/sql/Raster-operators.md
               - Raster map algebra: api/sql/Raster-map-algebra.md

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

@@ -22,12 +22,14 @@ import org.apache.spark.sql.catalyst.FunctionIdentifier
 import org.apache.spark.sql.catalyst.analysis.FunctionRegistry.FunctionBuilder
 import org.apache.spark.sql.catalyst.expressions.{ExpectsInputTypes, Expression, ExpressionInfo, Literal}
 import org.apache.spark.sql.expressions.Aggregator
-import org.apache.spark.sql.sedona_sql.expressions.collect.{ST_Collect}
+import org.apache.spark.sql.sedona_sql.expressions.collect.ST_Collect
 import org.apache.spark.sql.sedona_sql.expressions.raster._
 import org.apache.spark.sql.sedona_sql.expressions._
+import org.geotools.coverage.grid.GridCoverage2D
 import org.locationtech.jts.geom.Geometry
 import org.locationtech.jts.operation.buffer.BufferParameters
 
+import scala.collection.mutable.ArrayBuffer
 import scala.reflect.ClassTag
 
 object Catalog {
@@ -259,6 +261,8 @@ object Catalog {
     function[RS_Resample]()
   )
 
+  val rasterAggregateExpression: Aggregator[(GridCoverage2D, Int), ArrayBuffer[BandData], GridCoverage2D] = new RS_Union_Aggr
+
   val aggregateExpressions: Seq[Aggregator[Geometry, Geometry, Geometry]] = Seq(
     new ST_Union_Aggr,
     new ST_Envelope_Aggr,

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

@@ -37,6 +37,7 @@ object UdfRegistrator {
     }
 Catalog.aggregateExpressions.foreach(f => sparkSession.udf.register(f.getClass.getSimpleName, functions.udaf(f))) // SPARK3 anchor
 //Catalog.aggregateExpressions_UDAF.foreach(f => sparkSession.udf.register(f.getClass.getSimpleName, f)) // SPARK2 anchor
+    sparkSession.udf.register(Catalog.rasterAggregateExpression.getClass.getSimpleName, functions.udaf(Catalog.rasterAggregateExpression))
   }
 
   def dropAll(sparkSession: SparkSession): Unit = {
@@ -45,5 +46,6 @@ Catalog.aggregateExpressions.foreach(f => sparkSession.udf.register(f.getClass.g
     }
 Catalog.aggregateExpressions.foreach(f => sparkSession.sessionState.functionRegistry.dropFunction(FunctionIdentifier(f.getClass.getSimpleName))) // SPARK3 anchor
 //Catalog.aggregateExpressions_UDAF.foreach(f => sparkSession.sessionState.functionRegistry.dropFunction(FunctionIdentifier(f.getClass.getSimpleName))) // SPARK2 anchor
+    sparkSession.sessionState.functionRegistry.dropFunction(FunctionIdentifier(Catalog.rasterAggregateExpression.getClass.getSimpleName))
   }
 }

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

@@ -0,0 +1,130 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *   http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ * KIND, either express or implied.  See the License for the
+ * specific language governing permissions and limitations
+ * under the License.
+ */
+
+package org.apache.spark.sql.sedona_sql.expressions.raster
+
+import org.apache.sedona.common.raster.{RasterAccessors, RasterBandAccessors}
+import org.apache.sedona.common.utils.RasterUtils
+import org.apache.spark.sql.Encoder
+import org.apache.spark.sql.catalyst.encoders.ExpressionEncoder
+import org.apache.spark.sql.expressions.Aggregator
+import org.geotools.coverage.GridSampleDimension
+import org.geotools.coverage.grid.GridCoverage2D
+
+import java.awt.image.WritableRaster
+import javax.media.jai.RasterFactory
+import scala.collection.mutable
+import scala.collection.mutable.ArrayBuffer
+
+case class BandData(var bandInt: Array[Int], var bandDouble: Array[Double], var index: Int, var isIntegral: Boolean)
+
+/**
+ * Return a raster containing bands at given indexes from all rasters in a given column
+ */
+class RS_Union_Aggr extends Aggregator[(GridCoverage2D, Int), ArrayBuffer[BandData], GridCoverage2D]  {
+
+  var width: Int = -1
+
+  var height: Int = -1
+
+  var referenceRaster: GridCoverage2D = _
+
+  var gridSampleDimension: mutable.Map[Int, GridSampleDimension] = new mutable.HashMap()
+
+  def zero: ArrayBuffer[BandData] = ArrayBuffer[BandData]()
+
+  /**
+   * Valid raster shape to be the same in the given column
+   */
+  def checkRasterShape(raster: GridCoverage2D): Boolean = {
+    // first iteration
+    if (width == -1 && height == -1) {
+      width = RasterAccessors.getWidth(raster)
+      height = RasterAccessors.getHeight(raster)
+      referenceRaster = raster
+      true
+    } else {
+      val widthNewRaster = RasterAccessors.getWidth(raster)
+      val heightNewRaster = RasterAccessors.getHeight(raster)
+
+      width == widthNewRaster && height == heightNewRaster
+    }
+  }
+
+  def reduce(buffer: ArrayBuffer[BandData], input: (GridCoverage2D, Int)): ArrayBuffer[BandData] = {
+    val raster = input._1
+    if (!checkRasterShape(raster)) {
+      throw new IllegalArgumentException("Rasters provides should be of the same shape.")
+    }
+    if (gridSampleDimension.contains(input._2)) {
+      throw new IllegalArgumentException("Indexes shouldn't be repeated. Index should be in an arithmetic sequence.")
+    }
+
+    val rasterData = RasterUtils.getRaster(raster.getRenderedImage)
+    val isIntegral = RasterUtils.isDataTypeIntegral(rasterData.getDataBuffer.getDataType)
+
+    val bandData = if (isIntegral) {
+      val band = rasterData.getSamples(0, 0, width, height, 0, null.asInstanceOf[Array[Int]])
+      BandData(band, null, input._2, isIntegral)
+    } else {
+      val band = rasterData.getSamples(0, 0, width, height, 0, null.asInstanceOf[Array[Double]])
+      BandData(null, band, input._2, isIntegral)
+    }
+    gridSampleDimension = gridSampleDimension + (input._2 -> raster.getSampleDimension(0))
+
+    buffer += bandData
+  }
+
+  def merge(buffer1: ArrayBuffer[BandData], buffer2: ArrayBuffer[BandData]): ArrayBuffer[BandData] = {
+    ArrayBuffer.concat(buffer1, buffer2)
+  }
+
+  def finish(merged: ArrayBuffer[BandData]): GridCoverage2D = {
+    val sortedMerged = merged.sortBy(_.index)
+    val numBands = sortedMerged.length
+    val rasterData = RasterUtils.getRaster(referenceRaster.getRenderedImage)
+    val dataTypeCode = rasterData.getDataBuffer.getDataType
+    val resultRaster: WritableRaster = RasterFactory.createBandedRaster(dataTypeCode, width, height, numBands, null)
+    val gridSampleDimensions: Array[GridSampleDimension] = new Array[GridSampleDimension](numBands)
+    var indexCheck = 1
+
+    for (bandData: BandData <- sortedMerged) {
+      if (bandData.index != indexCheck) {
+        throw new IllegalArgumentException("Indexes should be in a valid arithmetic sequence.")
+      }
+      indexCheck += 1
+      gridSampleDimensions(bandData.index - 1) = gridSampleDimension(bandData.index)
+      if(RasterUtils.isDataTypeIntegral(dataTypeCode))
+        resultRaster.setSamples(0, 0, width, height, (bandData.index - 1), bandData.bandInt)
+      else
+        resultRaster.setSamples(0, 0, width, height, bandData.index - 1, bandData.bandDouble)
+
+    }
+    val noDataValue = RasterBandAccessors.getBandNoDataValue(referenceRaster)
+    RasterUtils.clone(resultRaster, referenceRaster.getGridGeometry, gridSampleDimensions, referenceRaster, noDataValue, true)
+  }
+
+  val serde = ExpressionEncoder[GridCoverage2D]
+
+  val bufferSerde = ExpressionEncoder[ArrayBuffer[BandData]]
+
+  def outputEncoder: ExpressionEncoder[GridCoverage2D] = serde
+
+  def bufferEncoder: Encoder[ArrayBuffer[BandData]] = bufferSerde
+}

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

@@ -20,11 +20,12 @@ package org.apache.sedona.sql
 
 import org.apache.sedona.common.raster.MapAlgebra
 import org.apache.sedona.common.utils.RasterUtils
+import org.apache.spark.sql.expressions.Window
 import org.apache.spark.sql.{Row, SaveMode}
-import org.apache.spark.sql.functions.{col, collect_list, expr}
+import org.apache.spark.sql.functions.{col, collect_list, expr, row_number}
 import org.geotools.coverage.grid.GridCoverage2D
 import org.junit.Assert.{assertEquals, assertNull, assertTrue}
-import org.locationtech.jts.geom.{Coordinate, Geometry, Point}
+import org.locationtech.jts.geom.{Coordinate, Geometry}
 import org.scalatest.{BeforeAndAfter, GivenWhenThen}
 
 import java.awt.image.DataBuffer
@@ -879,6 +880,28 @@ class rasteralgebraTest extends TestBaseScala with BeforeAndAfter with GivenWhen
       assertEquals(expected, actual)
     }
 
+    it("Passed RS_Union_Aggr") {
+      var df = sparkSession.read.format("binaryFile").load(resourceFolder + "raster/test1.tiff")
+        .union(sparkSession.read.format("binaryFile").load(resourceFolder + "raster/test1.tiff")
+          .union(sparkSession.read.format("binaryFile").load(resourceFolder + "raster/test1.tiff")))
+        .withColumn("raster", expr("RS_FromGeoTiff(content) as raster"))
+        .withColumn("index", row_number().over(Window.orderBy("raster")))
+        .selectExpr("raster", "index")
+
+      val dfTest = sparkSession.read.format("binaryFile").load(resourceFolder + "raster/test1.tiff")
+        .selectExpr("RS_FromGeoTiff(content) as raster")
+
+      df = df.selectExpr("RS_Union_aggr(raster, index) as rasters")
+
+      val actualBands = df.selectExpr("RS_NumBands(rasters)").first().get(0)
+      val expectedBands = 3
+      assertEquals(expectedBands, actualBands)
+
+      val actualMetadata = df.selectExpr("RS_Metadata(rasters)").first().getSeq(0).slice(0, 9)
+      val expectedMetadata = dfTest.selectExpr("RS_Metadata(raster)").first().getSeq(0).slice(0, 9)
+      assertTrue(expectedMetadata.equals(actualMetadata))
+    }
+
     it("Passed RS_ZonalStats") {
       var df = sparkSession.read.format("binaryFile").load(resourceFolder + "raster_geotiff_color/FAA_UTM18N_NAD83.tif")
       df = df.selectExpr("RS_FromGeoTiff(content) as raster", "ST_GeomFromWKT('POLYGON ((236722 4204770, 243900 4204770, 243900 4197590, 221170 4197590, 236722 4204770))', 26918) as geom")