Parquet Export¶
Module: spectralbridge.parquet_export.
Parquet export utilities for ENVI reflectance products.
The helpers in this module are intentionally lightweight at import time so that
unit tests can import :mod:spectralbridge.parquet_export without requiring
heavy geospatial dependencies. Expensive libraries (rasterio, pandas,
pyarrow) are imported lazily inside :func:build_parquet_from_envi.
API glossary ยท Workflow / scientific context
Signatures and documentation below are collected from the current source. Advanced objects are not newly declared stable by inclusion here.
Audience: advanced.
read_envi_in_chunks ¶
read_envi_in_chunks(envi_img: Path, envi_hdr: Path, parquet_name: str, *, chunk_size: int = 50000)
Yield DataFrame chunks from an ENVI cube along with shared metadata.
Audience: advanced.
parquet_exists_and_valid ¶
parquet_exists_and_valid(parquet_path: Path) -> bool
Return True when parquet_path exists and has a readable schema.
Audience: advanced.
build_parquet_from_envi ¶
build_parquet_from_envi(envi_img: Path, envi_hdr: Path, parquet_path: Path, chunk_size: int = 50000, *, num_cpus: int | None = None) -> None
Read an ENVI cube and emit a wide per-pixel Parquet table.
Audience: advanced.
ensure_parquet_from_envi ¶
ensure_parquet_from_envi(envi_img: Path, envi_hdr: Path, parquet_path: Path, *, chunk_size: int = 50000, num_cpus: int | None = None) -> Path
Reuse a readable Parquet schema or rebuild the table from the existing ENVI image/header in bounded chunks.
Audience: advanced.
repair_lonlat_in_place ¶
repair_lonlat_in_place(folder: Path)
Add missing coordinate columns to eligible Parquets in place using matching ENVI geometry.
Audience: advanced.
ensure_parquet_for_envi ¶
ensure_parquet_for_envi(envi_img: Path, logger, *, chunk_size: int = 2048, ray_cpus: int | None = None) -> Path | None
Ensure a .parquet sidecar exists for envi_img.
This treats the _undarkened_envi.img/.hdr convolution-only products the
same as their brightness-adjusted counterparts so QA comparisons stay
aligned.