# Xarray-binfile ```{centered} Read and write raw binary files using the familiar interface from the Xarray library. ``` Xarray-binfile provides an xarray backend for raw binary files that are not self-describing on their own. You supply the metadata needed to interpret each file, and xarray-binfile handles lazy loading, chunked parallel computation with Dask, and writing results back to raw binary when you need to interoperate with an existing file layout. ::::{grid} 1 1 2 3 :class-container: text-center :gutter: 3 :::{grid-item-card} :link: getting-started/about :link-type: doc :class-header: bg-light About The Backend ^^^ Learn how xarray-binfile maps raw binary files into labeled xarray datasets and data arrays. ::: :::{grid-item-card} :link: getting-started/installation :link-type: doc :class-header: bg-light Install And Configure ^^^ Install the package, understand the runtime dependencies, and get ready to open binary datasets lazily. ::: :::{grid-item-card} :link: tutorials/read :link-type: doc :class-header: bg-light Read Lazily With Dask ^^^ Open binary files with chunking, inspect the Dask graph, and compute only the results you need. ::: :::{grid-item-card} :link: tutorials/write :link-type: doc :class-header: bg-light Write Derived Results ^^^ Take a computed xarray result and write it back to raw binary files when you need to match an external format. ::: :::{grid-item-card} :link: tutorials/parallel-and-larger-than-memory :link-type: doc :class-header: bg-light Parallel And Out-Of-Core ^^^ Open files in parallel, pick a Dask scheduler, and process datasets larger than memory by streaming chunks. ::: :::{grid-item-card} :link: tutorials/benchmarking :link-type: doc :class-header: bg-light Benchmark Your Data ^^^ Measure read and compute performance on your own files, and see how results shift with dtype, dimension size, and hardware. ::: :::{grid-item-card} :link: tutorials/examples :link-type: doc :class-header: bg-light Practical Examples ^^^ Batch-convert dtypes across files, extract a probe time series to CSV, and draw quick matplotlib plots. ::: :::{grid-item-card} :link: references/api-reference :link-type: doc :class-header: bg-light API Reference ^^^ See the public modules, metadata protocols, backend entrypoint, and tutorial helpers. ::: :::{grid-item-card} :link: references/further-reading :link-type: doc :class-header: bg-light Further Reading ^^^ Go deeper on xarray, Dask, NumPy binary I/O, and related simulation workflows. ::: :::: ```{important} Raw binary files usually need external metadata such as shape, coordinate values, dtype, and filename conventions. Xarray-binfile does not try to infer that information magically. Instead, it asks you to provide it explicitly so the backend can remain predictable and interoperable. ``` ```{warning} Raw binary data is machine-specific unless you control the full convention explicitly. Byte order differences (little-endian vs big-endian), numeric representation, and layout assumptions may break portability across architectures. Validate dtype and endianness before using files generated on another machine. ``` ```{note} For portable archival or data exchange workflows, prefer standard xarray-supported formats such as NetCDF or Zarr. Raw binary output is most useful when you need to read or write an existing binary convention used by a simulation code or downstream tool. ```