"Gentle Introduction to Arrow / Record Batches" #11336

docs/source/index.rst

@@ -118,6 +118,7 @@ To get started, see
    user-guide/crate-configuration
    user-guide/cli/index
    user-guide/dataframe
+   user-guide/arrow-introduction
    user-guide/expressions
    user-guide/sql/index
    user-guide/configs

docs/source/user-guide/arrow-introduction.md

@@ -0,0 +1,291 @@
+<!---
+  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.
+-->
+
+# Introduction to `Arrow` & RecordBatches
+
+```{contents}
+:local:
+:depth: 2
+```
+
+This guide helps DataFusion users understand [Apache Arrow]—a language-independent, in-memory columnar format and development platform for analytics. It defines a standardized columnar representation that enables different systems and languages (e.g., Rust and Python) to share data with zero-copy interchange, avoiding serialization overhead. A core building block is the [`RecordBatch`] format. While you may never need to work with Arrow directly, this knowledge becomes valuable when using DataFusion's extension points or debugging performance issues.
+
+**Why Arrow is central to DataFusion**: Arrow provides the unified type system that makes DataFusion possible. When you query a CSV file, join it with a Parquet file, and aggregate results from JSON—it all works seamlessly because every data source is converted to Arrow's common representation. This unified type system, combined with Arrow's columnar format, enables DataFusion to execute efficient vectorized operations across any combination of data sources while benefiting from zero-copy data sharing between query operators.
+
+## Why Columnar? The Arrow Advantage
+
+Apache Arrow is an open **specification** that defines a common way to organize analytical data in memory. Think of it as a set of best practices that different systems agree to follow, not a database or programming language.
+
+### Row-oriented vs Columnar Layout
+
+Quick visual: row-major (left) vs Arrow's columnar layout (right). For a deeper primer, see the [arrow2 guide].
+
+```
+Traditional Row Storage:          Arrow Columnar Storage:
+┌──────────────────┐              ┌─────────┬─────────┬──────────┐
+│ id │ name │ age  │              │   id    │  name   │   age    │
+├────┼──────┼──────┤              ├─────────┼─────────┼──────────┤
+│ 1  │  A   │  30  │              │ [1,2,3] │ [A,B,C] │[30,25,35]│
+│ 2  │  B   │  25  │              └─────────┴─────────┴──────────┘
+│ 3  │  C   │  35  │                   ↑          ↑         ↑
+└──────────────────┘              Int32Array StringArray Int32Array
+(read entire rows)                (process entire columns at once)
+```
+
+### Why This Matters
+
+- **Unified Type System**: All data sources (CSV, Parquet, JSON) convert to the same Arrow types, enabling seamless cross-format queries
+- **Vectorized Execution**: Process entire columns at once using SIMD instructions
+- **Cache Efficiency**: Scanning specific columns doesn't load unnecessary data into CPU cache
+- **Zero-Copy Data Sharing**: Systems can share Arrow data without conversion overhead
+
+Arrow is widely adopted for in-memory analytics precisely because of its **columnar format**—systems that natively store or process data in Arrow (DataFusion, Polars, InfluxDB 3.0), and runtimes that convert to Arrow for interchange (DuckDB, Spark, pandas), all organize data by column rather than by row. This cross-language, cross-platform adoption of the columnar model enables seamless data flow between systems with minimal conversion overhead.
+
+Within this columnar design, Arrow's standard unit for packaging data is the **RecordBatch**—the key to making columnar format practical for real-world query engines.
+
+## What is a RecordBatch? (And Why Batch?)
+
+A **[`RecordBatch`]** represents a horizontal slice of a table—a collection of equal-length columnar arrays that conform to a defined schema. Each column within the slice is a contiguous Arrow array, and all columns have the same number of rows (length). This chunked, immutable unit enables efficient streaming and parallel execution.
+
+Think of it as having two perspectives:
+
+- **Columnar inside**: Each column (`id`, `name`, `age`) is a contiguous array optimized for vectorized operations
+- **Row-chunked externally**: The batch represents a chunk of rows (e.g., rows 1-1000), making it a manageable unit for streaming
+
+RecordBatches are **immutable snapshots**—once created, they cannot be modified. Any transformation produces a _new_ RecordBatch, enabling safe parallel processing without locks or coordination overhead.
+
+This design allows DataFusion to process streams of row-based chunks while gaining maximum performance from the columnar layout. Let's see how this works in practice.
+
+### Configuring Batch Size
+
+DataFusion uses a default batch size of 8192 rows per RecordBatch, balancing memory efficiency with vectorization benefits. You can adjust this via the session configuration or the [`datafusion.execution.batch_size`] configuration setting:
+
+```rust
+use datafusion::execution::config::SessionConfig;
+use datafusion::prelude::*;
+
+let config = SessionConfig::new().with_batch_size(8192); // default value
+let ctx = SessionContext::with_config(config);
+```
+
+You can also query and modify this setting using SQL:
+
+```sql
+SHOW datafusion.execution.batch_size;
+SET datafusion.execution.batch_size TO 1024;
+```
+
+See [configuration settings] for more details.
+
+## From files to Arrow
+
+When you call [`read_csv`], [`read_parquet`], [`read_json`] or [`read_avro`], DataFusion decodes those formats into Arrow arrays and streams them to operators as RecordBatches.
+
+The example below shows how to read data from different file formats. Each `read_*` method returns a [`DataFrame`] that represents a query plan. When you call [`.collect()`], DataFusion executes the plan and returns results as a `Vec<RecordBatch>`—the actual columnar data in Arrow format.
+
+```rust
+use datafusion::prelude::*;
+
+#[tokio::main]
+async fn main() -> datafusion::error::Result<()> {
+    let ctx = SessionContext::new();
+
+    // Pick ONE of these per run (each returns a new DataFrame):
+    let df = ctx.read_csv("data.csv", CsvReadOptions::new()).await?;
+    // let df = ctx.read_parquet("data.parquet", ParquetReadOptions::default()).await?;
+    // let df = ctx.read_json("data.ndjson", NdJsonReadOptions::default()).await?; // requires "json" feature; expects newline-delimited JSON (NDJSON)
+    // let df = ctx.read_avro("data.avro", AvroReadOptions::default()).await?;     // requires "avro" feature
+
+    let batches = df
+        .select(vec![col("id")])?
+        .filter(col("id").gt(lit(10)))?
+        .collect()
+        .await?; // Vec<RecordBatch>
+
+    Ok(())
+}
+```
+
+## Streaming Through the Engine
+
+DataFusion processes queries as pull-based pipelines where operators request batches from their inputs. This streaming approach enables early result production, bounds memory usage (spilling to disk only when necessary), and naturally supports parallel execution across multiple CPU cores.
+
+```
+A user's query: SELECT name FROM 'data.parquet' WHERE id > 10
+
+The DataFusion Pipeline:
+┌─────────────┐    ┌──────────────┐    ┌────────────────┐    ┌──────────────────┐    ┌──────────┐
+│ Parquet     │───▶│ Scan         │───▶│ Filter         │───▶│ Projection       │───▶│ Results  │
+│ File        │    │ Operator     │    │ Operator       │    │ Operator         │    │          │
+└─────────────┘    └──────────────┘    └────────────────┘    └──────────────────┘    └──────────┘
+                   (reads data)        (id > 10)             (keeps "name" col)
+                   RecordBatch ───▶    RecordBatch ────▶     RecordBatch ────▶        RecordBatch
+```
+
+In this pipeline, [`RecordBatch`]es are the "packages" of columnar data that flow between the different stages of query execution. Each operator processes batches incrementally, enabling the system to produce results before reading the entire input.
+
+## Minimal: build a RecordBatch in Rust
+
+Sometimes you need to create Arrow data programmatically rather than reading from files. This example shows the core building blocks: creating typed arrays (like [`Int32Array`] for numbers), defining a [`Schema`] that describes your columns, and assembling them into a [`RecordBatch`].
+
+Note: You'll see [`Arc`] used frequently in the code—Arrow arrays are wrapped in `Arc` (atomically reference-counted pointers) to enable cheap, thread-safe sharing across operators and tasks. [`ArrayRef`] is simply a type alias for `Arc<dyn Array>`.
+
+```rust
+use std::sync::Arc;
+use arrow_schema::ArrowError;
+use arrow_array::{ArrayRef, Int32Array, StringArray, RecordBatch};
+use arrow_schema::{DataType, Field, Schema};
+
+fn make_batch() -> Result<RecordBatch, ArrowError> {
+    let ids = Int32Array::from(vec![1, 2, 3]);
+    let names = StringArray::from(vec![Some("alice"), None, Some("carol")]);
+
+    let schema = Arc::new(Schema::new(vec![
+        Field::new("id", DataType::Int32, false),
+        Field::new("name", DataType::Utf8, true),
+    ]));
+
+    let cols: Vec<ArrayRef> = vec![Arc::new(ids), Arc::new(names)];
+    RecordBatch::try_new(schema, cols)
+}
+```
+
+## Query an in-memory batch with DataFusion
+
+Once you have a [`RecordBatch`], you can query it with DataFusion using a [`MemTable`]. This is useful for testing, processing data from external systems, or combining in-memory data with other sources. The example below creates a batch, wraps it in a [`MemTable`], registers it as a named table, and queries it using SQL—demonstrating how Arrow serves as the bridge between your data and DataFusion's query engine.
+
+```rust
+use std::sync::Arc;
+use arrow_array::{Int32Array, StringArray, RecordBatch};
+use arrow_schema::{DataType, Field, Schema};
+use datafusion::datasource::MemTable;
+use datafusion::prelude::*;
+
+#[tokio::main]
+async fn main() -> datafusion::error::Result<()> {
+    let ctx = SessionContext::new();
+
+    // build a batch
+    let schema = Arc::new(Schema::new(vec![
+        Field::new("id", DataType::Int32, false),
+        Field::new("name", DataType::Utf8, true),
+    ]));
+    let batch = RecordBatch::try_new(
+        schema.clone(),
+        vec![
+            Arc::new(Int32Array::from(vec![1, 2, 3])) as _,
+            Arc::new(StringArray::from(vec![Some("foo"), Some("bar"), None])) as _,
+        ],
+    )?;
+
+    // expose it as a table
+    let table = MemTable::try_new(schema, vec![vec![batch]])?;
+    ctx.register_table("people", Arc::new(table))?;
+
+    // query it
+    let df = ctx.sql("SELECT id, upper(name) AS name FROM people WHERE id >= 2").await?;
+    df.show().await?;
+    Ok(())
+}
+```
+
+## Common Pitfalls
+
+When working with Arrow and RecordBatches, watch out for these common issues:
+
+- **Schema consistency**: All batches in a stream must share the exact same [`Schema`]. For example, you can't have one batch where a column is [`Int32`] and the next where it's [`Int64`], even if the values would fit
+- **Immutability**: Arrays are immutable—to "modify" data, you must build new arrays or new RecordBatches. For instance, to change a value in an array, you'd create a new array with the updated value
+- **Row by Row Processing**: Avoid iterating over Arrays element by element when possible, and use Arrow's built-in compute functions instead
+- **Type mismatches**: Mixed input types across files may require explicit casts. For example, a string column `"123"` from a CSV file won't automatically join with an integer column `123` from a Parquet file—you'll need to cast one to match the other. Use Arrow's [`cast`] kernel where appropriate
+- **Batch size assumptions**: Don't assume a particular batch size; always iterate until the stream ends. One file might produce 8192-row batches while another produces 1024-row batches
+
+## When Arrow knowledge is needed (Extension Points)
+
+For many use cases, you don't need to know about Arrow. DataFusion handles the conversion from formats like CSV and Parquet for you. However, Arrow becomes important when you use DataFusion's **[extension points]** to add your own custom functionality.
+
+These APIs are where you can plug your own code into the engine, and they often operate directly on Arrow [`RecordBatch`] streams.
+
+- **[`TableProvider`] (Custom Data Sources)**: This is the most common extension point. You can teach DataFusion how to read from any source—a custom file format, a network API, a different database—by implementing the [`TableProvider`] trait. Your implementation will be responsible for creating [`RecordBatch`]es to stream data into the engine. See the [Custom Table Providers guide] for detailed examples.
+
+- **[User-Defined Functions (UDFs)]**: If you need to perform a custom transformation on your data that isn't built into DataFusion, you can write a UDF. Your function will receive data as Arrow arrays (inside a [`RecordBatch`]) and must produce an Arrow array as its output.
+
+- **[Custom Optimizer Rules and Physical Operators]**: For advanced use cases, you can add your own optimizer rules or implement entirely new physical operators by implementing the [`ExecutionPlan`] trait. While optimizer rules primarily reason about schemas and plan properties, ExecutionPlan implementations directly produce and consume streams of [`RecordBatch`]es during query execution.
+
+In short, knowing Arrow is key to unlocking the full power of DataFusion's modular and extensible architecture.
+
+## Further reading
+
+**Arrow Documentation:**
+
+- [Arrow Format Introduction](https://arrow.apache.org/docs/format/Intro.html) - Understand the Arrow specification and why it enables zero-copy data sharing
+- [Arrow Columnar Format](https://arrow.apache.org/docs/format/Columnar.html) - Deep dive into memory layout for performance optimization
+- [Arrow Rust Documentation](https://docs.rs/arrow/latest/arrow/) - Complete API reference for the Rust implementation
+
+**DataFusion Extension Guides:**
+
+- [Custom Table Providers](../library-user-guide/custom-table-providers.md) - Build custom data sources that stream RecordBatches
+- [Adding UDFs](../library-user-guide/functions/adding-udfs.md) - Create efficient user-defined functions working with Arrow arrays
+- [Extending Operators](../library-user-guide/extending-operators.md) - Implement custom ExecutionPlan operators for specialized processing
+
+**Key API References:**
+
+- [RecordBatch](https://docs.rs/arrow-array/latest/arrow_array/struct.RecordBatch.html) - The fundamental data structure for streaming columnar data
+- [ExecutionPlan](https://docs.rs/datafusion/latest/datafusion/physical_plan/trait.ExecutionPlan.html) - Trait for implementing custom physical operators
+- [TableProvider](https://docs.rs/datafusion/latest/datafusion/datasource/trait.TableProvider.html) - Interface for custom data sources
+- [ScalarUDF](https://docs.rs/datafusion/latest/datafusion/logical_expr/struct.ScalarUDF.html) - Building scalar user-defined functions
+- [MemTable](https://docs.rs/datafusion/latest/datafusion/datasource/struct.MemTable.html) - Working with in-memory Arrow data
+
+**Academic Paper:**
+
+- [Apache Arrow DataFusion: A Fast, Embeddable, Modular Analytic Query Engine](https://dl.acm.org/doi/10.1145/3626246.3653368) - Architecture and design decisions (SIGMOD 2024)
+
+[apache arrow]: https://arrow.apache.org/docs/index.html
+[`arc`]: https://doc.rust-lang.org/std/sync/struct.Arc.html
+[`arrayref`]: https://docs.rs/arrow-array/latest/arrow_array/array/type.ArrayRef.html
+[`cast`]: https://docs.rs/arrow/latest/arrow/compute/fn.cast.html
+[`field`]: https://docs.rs/arrow-schema/latest/arrow_schema/struct.Field.html
+[`schema`]: https://docs.rs/arrow-schema/latest/arrow_schema/struct.Schema.html
+[`datatype`]: https://docs.rs/arrow-schema/latest/arrow_schema/enum.DataType.html
+[`int32array`]: https://docs.rs/arrow-array/latest/arrow_array/array/struct.Int32Array.html
+[`stringarray`]: https://docs.rs/arrow-array/latest/arrow_array/array/struct.StringArray.html
+[`int32`]: https://docs.rs/arrow-schema/latest/arrow_schema/enum.DataType.html#variant.Int32
+[`int64`]: https://docs.rs/arrow-schema/latest/arrow_schema/enum.DataType.html#variant.Int64
+[extension points]: ../library-user-guide/extensions.md
+[`tableprovider`]: https://docs.rs/datafusion/latest/datafusion/datasource/trait.TableProvider.html
+[custom table providers guide]: ../library-user-guide/custom-table-providers.md
+[user-defined functions (udfs)]: ../library-user-guide/functions/adding-udfs.md
+[custom optimizer rules and physical operators]: ../library-user-guide/extending-operators.md
+[`executionplan`]: https://docs.rs/datafusion/latest/datafusion/physical_plan/trait.ExecutionPlan.html
+[`.register_table()`]: https://docs.rs/datafusion/latest/datafusion/execution/context/struct.SessionContext.html#method.register_table
+[`.sql()`]: https://docs.rs/datafusion/latest/datafusion/execution/context/struct.SessionContext.html#method.sql
+[`.show()`]: https://docs.rs/datafusion/latest/datafusion/dataframe/struct.DataFrame.html#method.show
+[`memtable`]: https://docs.rs/datafusion/latest/datafusion/datasource/struct.MemTable.html
+[`sessioncontext`]: https://docs.rs/datafusion/latest/datafusion/execution/context/struct.SessionContext.html
+[`csvreadoptions`]: https://docs.rs/datafusion/latest/datafusion/execution/options/struct.CsvReadOptions.html
+[`parquetreadoptions`]: https://docs.rs/datafusion/latest/datafusion/execution/options/struct.ParquetReadOptions.html
+[`recordbatch`]: https://docs.rs/arrow-array/latest/arrow_array/struct.RecordBatch.html
+[`read_csv`]: https://docs.rs/datafusion/latest/datafusion/execution/context/struct.SessionContext.html#method.read_csv
+[`read_parquet`]: https://docs.rs/datafusion/latest/datafusion/execution/context/struct.SessionContext.html#method.read_parquet
+[`read_json`]: https://docs.rs/datafusion/latest/datafusion/execution/context/struct.SessionContext.html#method.read_json
+[`read_avro`]: https://docs.rs/datafusion/latest/datafusion/execution/context/struct.SessionContext.html#method.read_avro
+[`dataframe`]: https://docs.rs/datafusion/latest/datafusion/dataframe/struct.DataFrame.html
+[`.collect()`]: https://docs.rs/datafusion/latest/datafusion/dataframe/struct.DataFrame.html#method.collect
+[arrow2 guide]: https://jorgecarleitao.github.io/arrow2/main/guide/arrow.html#what-is-apache-arrow
+[configuration settings]: configs.md
+[`datafusion.execution.batch_size`]: configs.md#setting-configuration-options

docs/source/user-guide/dataframe.md

@@ -19,6 +19,8 @@
 
 # DataFrame API
 
+## DataFrame overview
+
 A DataFrame represents a logical set of rows with the same named columns,
 similar to a [Pandas DataFrame] or [Spark DataFrame].