Kontinuation commented on code in PR #573: URL: https://github.com/apache/sedona-db/pull/573#discussion_r2780493637
########## rust/sedona-spatial-join/src/probe/knn_results_merger.rs: ########## @@ -0,0 +1,2226 @@ +// 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. + +use core::f64; +use std::ops::Range; +use std::sync::Arc; + +use arrow::array::{ + Array, ArrayBuilder, AsArray, Float64Array, ListArray, OffsetBufferBuilder, PrimitiveBuilder, + RecordBatch, StructArray, UInt64Array, +}; +use arrow::buffer::OffsetBuffer; +use arrow::compute::{concat, concat_batches, interleave}; +use arrow::datatypes::{DataType, Field, Float64Type, Schema, SchemaRef, UInt64Type}; +use arrow_array::ArrayRef; +use arrow_schema::Fields; +use datafusion::config::SpillCompression; +use datafusion_common::{arrow_datafusion_err, DataFusionError, Result}; +use datafusion_execution::disk_manager::RefCountedTempFile; +use datafusion_execution::runtime_env::RuntimeEnv; +use datafusion_physical_plan::metrics::SpillMetrics; +use sedona_common::sedona_internal_err; + +use crate::index::spatial_index::DISTANCE_TOLERANCE; +use crate::utils::spill::{RecordBatchSpillReader, RecordBatchSpillWriter}; + +/// [UnprocessedKNNResultBatch] represents the KNN results produced by probing the spatial index. +/// An [UnprocessedKNNResultBatch] may include KNN results for multiple probe rows. +/// +/// The KNN results are stored in a StructArray, where each row corresponds to a KNN result. +/// The results for the same probe row are stored in contiguous rows, and the offsets to +/// split the results into groups per probe row are stored in the `offsets` field. +/// +/// Each probe row has a unique index. The index must be strictly increasing +/// across probe rows. The sequence of index across the entire sequence of ingested +/// [UnprocessedKNNResultBatch] must also be strictly increasing. The index is computed based on +/// the 0-based index of the probe row in this probe partition. +/// +/// The KNN results are filtered, meaning that the original KNN results obtained by probing +/// the spatial index may be further filtered based on some predicates. It is also possible that +/// all the KNN results for a probe row are filtered out. However, we still need to keep track of the +/// distances of unfiltered results to correctly compute the top-K distances before filtering. This +/// is critical for correctly merging KNN results from multiple partitions. Review Comment: It can be any condition, not necessarily related to the distance. The writing can be a bit confusing. I have extended the comment to explain it. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
