faxarray

User-friendly interface for Météo-France FA files with xarray integration

A Python package that wraps EPyGrAM with a clean, xarray-like API for handling FA files. Provides easy plotting, tar archive support, and NetCDF conversion.

Features

• Easy plotting with .plot() methods (like xarray)
• Native xarray backend - use xr.open_dataset() directly on FA files
• Tar archive support - open FA files directly from .tar.gz archives
• Multi-file conversion - combine multiple FA files into one NetCDF
• Simple API - no complex initialization required
• CLI tool for quick operations

Installation

Important: This package requires Python 3.11.

# Create a new environment with Python 3.11
conda create -n faxarray_env python=3.11
conda activate faxarray_env

# Install faxarray
pip install git+https://github.com/Debasish-Mahapatra/faxarray

Quick Start

Open FA files with xarray

import xarray as xr
import faxarray  # Just import to register the backend

# Open FA file - auto-detects format
ds = xr.open_dataset('pfABOFABOF+0001')

# All xarray operations work
print(ds['SURFTEMPERATURE'].mean())
ds.to_netcdf('output.nc')

Open FA files from tar archives

import faxarray as fx

# Open directly from tar archive
ds = fx.open_tar('pf20130101.tar.gz', '/tmp/extract')

# Use xarray filtering (lazy loading - fast!)
temp = ds['SURFTEMPERATURE'].isel(time=0)
temp.plot()

ds.close()  # Cleanup temp files

See docs/exploring_tar_data.md for full documentation.

Convert multiple FA files to NetCDF

import faxarray as fx

# Combine multiple files with de-accumulation
ds = fx.open_mfdataset(
    'pf*+*',
    variables=['SURFPREC.EAU.CON'],  # Limit memory
    deaccumulate=['SURFPREC.EAU.CON'],
    output_file='output.nc'
)

Command Line Interface

# File information
faxarray info file.fa
faxarray info file.fa --list-vars

# Convert single file
faxarray convert input.fa output.nc

# Convert multiple files with de-accumulation
faxarray convert-multi 'pf*+*' output.nc \
    -v SURFPREC.EAU.CON \
    -d SURFPREC.EAU.CON

API Reference

Functions

Function	Description
fx.open_fa(path)	Open single FA file
fx.open_tar(tarpath, temp_dir)	Open FA files from tar archive
fx.open_mfdataset(pattern, ...)	Combine multiple FA files
xr.open_dataset(path)	xarray backend (auto-registered)

Key Parameters

open_tar:

• tarpath - Path to tar archive
• temp_dir - Extraction directory (required)
• variables - List of variables to load
• pattern - Glob pattern to filter files

open_mfdataset:

• paths - Glob pattern or list of files
• variables - Variables to load (saves memory)
• deaccumulate - Variables to de-accumulate
• output_file - Stream directly to NetCDF

Documentation

• Exploring Tar Data - Full guide for tar archives
• Memory Usage - Tips for large datasets

Requirements

• Python 3.11
• numpy, xarray, netCDF4, matplotlib
• epygram (v2.0.7+, auto-installed)
• h5py, cartopy

License

MIT