-
Notifications
You must be signed in to change notification settings - Fork 7
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge pull request #81 from compomics/add-parquet
Add Parquet reading and writing for efficient storage of PSM lists
- Loading branch information
Showing
16 changed files
with
426 additions
and
132 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,130 @@ | ||
| """ | ||
| Reader and writer for a simple, lossless psm_utils Parquet format. | ||
| Similar to the :py:mod:`psm_utils.io.tsv` module, this module provides a reader and writer | ||
| for :py:class:`~psm_utils.psm_list.PSMList` objects in a lossless manner. However, Parquet provides | ||
| better performance and storage efficiency compared to TSV, and is recommended for large datasets. | ||
| """ | ||
|
|
||
| from __future__ import annotations | ||
|
|
||
| from pathlib import Path | ||
| from typing import Union | ||
|
|
||
| import pyarrow as pa | ||
| import pyarrow.parquet as pq | ||
| from pydantic import ValidationError | ||
|
|
||
| from psm_utils.io._base_classes import ReaderBase, WriterBase | ||
| from psm_utils.io.exceptions import PSMUtilsIOException | ||
| from psm_utils.psm import PSM | ||
| from psm_utils.psm_list import PSMList | ||
|
|
||
|
|
||
| class ParquetReader(ReaderBase): | ||
| def __init__(self, path: Union[str, Path], *args, **kwargs): | ||
| """ | ||
| Reader for Parquet files. | ||
| Parameters | ||
| ---------- | ||
| path : Union[str, Path] | ||
| Path to the Parquet file. | ||
| """ | ||
| self.path = path | ||
|
|
||
| def __iter__(self): | ||
| with pq.ParquetFile(self.path) as reader: | ||
| for batch in reader.iter_batches(): | ||
| for row in batch.to_pylist(): | ||
| # Convert map columns (rendered as lists of tuples) to dictionaries | ||
| row["metadata"] = dict(row["metadata"] or {}) | ||
| row["provenance_data"] = dict(row["provenance_data"] or {}) | ||
| row["rescoring_features"] = dict(row["rescoring_features"] or {}) | ||
|
|
||
| # Convert to PSM object and yield | ||
| try: | ||
| yield PSM(**row) | ||
| except ValidationError as e: | ||
| raise PSMUtilsIOException(f"Error while parsing row {row}:\n{e}") | ||
|
|
||
|
|
||
| class ParquetWriter(WriterBase): | ||
| def __init__(self, path: Union[str, Path], chunk_size: int = 1e6, *args, **kwargs): | ||
| """ | ||
| Writer for Parquet files. | ||
| Parameters | ||
| ---------- | ||
| path : Union[str, Path] | ||
| Path to the Parquet file. | ||
| chunk_size : int | ||
| Number of PSMs to write in a single batch. Default is 1e6. | ||
| """ | ||
| self.path = path | ||
| self.chunk_size = chunk_size | ||
|
|
||
| self._writer = None | ||
| self._psm_cache = [] | ||
|
|
||
| def __enter__(self): | ||
| self._writer = pq.ParquetWriter(self.path, schema=SCHEMA) | ||
| return self | ||
|
|
||
| def __exit__(self, *args, **kwargs): | ||
| self._flush() | ||
| self._writer.close() | ||
|
|
||
| def write_psm(self, psm: PSM): | ||
| """Write a single PSM to the Parquet file.""" | ||
| self._psm_cache.append(self._psm_to_entry(psm)) | ||
| if len(self._psm_cache) > self.chunk_size: | ||
| self._flush() | ||
|
|
||
| def write_file(self, psm_list: PSMList): | ||
| """Write a list of PSMs to the Parquet file.""" | ||
| with self: | ||
| for psm in psm_list: | ||
| self.write_psm(psm) | ||
|
|
||
| @staticmethod | ||
| def _psm_to_entry(psm: PSM) -> dict: | ||
| """Convert a PSM object to a dictionary suitable for writing to Parquet.""" | ||
| psm_dict = dict(psm) | ||
| psm_dict["peptidoform"] = str(psm.peptidoform) | ||
| return psm_dict | ||
|
|
||
| def _flush(self): | ||
| """Write the cached PSMs to the Parquet file.""" | ||
| if not self._psm_cache: | ||
| return | ||
| table = pa.Table.from_pylist(self._psm_cache, schema=SCHEMA) | ||
| self._writer.write_table(table) | ||
| self._psm_cache = [] | ||
|
|
||
|
|
||
| SCHEMA = pa.schema( | ||
| [ | ||
| ("peptidoform", pa.string()), | ||
| ("spectrum_id", pa.string()), | ||
| ("run", pa.string()), | ||
| ("collection", pa.string()), | ||
| ("spectrum", pa.string()), | ||
| ("is_decoy", pa.bool_()), | ||
| ("score", pa.float32()), | ||
| ("qvalue", pa.float32()), | ||
| ("pep", pa.float32()), | ||
| ("precursor_mz", pa.float32()), | ||
| ("retention_time", pa.float32()), | ||
| ("ion_mobility", pa.float32()), | ||
| ("protein_list", pa.list_(pa.string())), | ||
| ("rank", pa.int32()), | ||
| ("source", pa.string()), | ||
| ("provenance_data", pa.map_(pa.string(), pa.string())), | ||
| ("metadata", pa.map_(pa.string(), pa.string())), | ||
| ("rescoring_features", pa.map_(pa.string(), pa.float32())), | ||
| ] | ||
| ) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.