repository·main·Indexed 23 days ago
https://github.com/unidata/metpyA Python library for meteorological data analysis, providing tools for reading, calculating, and visualizing weather data. It integrates with the scientific Python stack, including numpy, scipy, matplotlib, pandas, pint, and xarray. The library includes the metpy.calc module for dry and moist thermodynamics, sounding and stability indices, dynamic and kinematic calculations, and boundary layer turbulence, as well as plotting utilities via metpy.plots.ctables and xarray integration through custom accessors.
MetPy is a collection of Python tools designed for reading, visualizing, and performing calculations with weather data. It is designed to integrate seamlessly with the scientific Python ecosystem, specifically numpy, scipy, and matplotlib.
MetPy follows semantic versioning, meaning 1.x releases are backwards compatible with previous 1.y releases. It supports Python version 3.11 and above.
MetPy aims to provide domain-specific tools for building meteorological and atmospheric science applications in Python. Its primary objectives include:
matplotlib and xarray whenever possible to maintain a focused, domain-specific toolset.The MetPy API is organized into several functional modules. You can access core functionality through the following submodules:
constants: Physical constants used in meteorological calculations.units: Tools for handling quantities with physical units (integration with Pint).io: Input/Output utilities for reading and writing meteorological data.remote: Tools for accessing data from remote sources.calc: A collection of meteorological calculation routines.plots: Functions for creating meteorological plots.plots.ctables: Color table definitions for plotting.interpolate: Interpolation routines for meteorological data.xarray: Integration and utilities for working with xarray datasets.metpy.plots.ctables for managing and accessing colortables, which are used to map data values to colors in meteorological plots. The module provides functions for reading and converting colortables, as well as a registry for managing them.MetPy provides integration with xarray through custom accessors. Once MetPy is installed and your Xarray objects are loaded, you can access MetPy-specific functionality directly on DataArray and Dataset objects using the .metpy attribute.
MetPyDataArrayAccessor: Provides meteorological calculations and utilities specifically for xarray.DataArray objects.MetPyDatasetAccessor: Provides meteorological calculations and utilities specifically for xarray.Dataset objects.These accessors allow you to perform complex meteorological operations while preserving the metadata and coordinate structures inherent to Xarray.
MetPy follows semantic versioning. For the 1.x release series, any 1.x release is guaranteed to be backwards compatible with an earlier 1.y release.
Backward compatibility definition:
Correct code that works on a 1.y version will work on a future 1.x version. Note that bug fixes may change behavior or cause previously incorrect code to stop working.
Version numbering convention:
1.x.y where x is incremented for new features.1.x.y where y is incremented for bug fixes only.2.0).When breaking changes are planned for the 1.x series, they will be preceded by MetpyDeprecationWarning or FutureWarning to allow users to migrate.
metpy.calc module have been updated to support xarray.DataArray for both inputs and outputs. Previously, these functions often returned pint.Quantity objects even when an xarray.DataArray was provided as input. Now, MetPy will attempt to return an xarray.DataArray when provided one, preserving xarray metadata and structure where possible. For detailed usage, refer to the MetPy xarray tutorial.MetPy is developing a simplified, declarative plotting interface inspired by the simplicity of GEMPAK. Instead of complex matplotlib configurations, the goal is to allow users to create meteorological plots by setting variables.
Planned and existing support includes:
Note: This interface is intended for common meteorological use cases and does not aim to replicate the full feature set of matplotlib.
The MetPy CLA is a license agreement, not a copyright assignment. Key points for contributors include:
MetPy relies on xarray for gridded data models and pint for robust unit handling. A key architectural goal is to move toward native xarray support to enable coordinate-aware calculations (e.g., vorticity, isentropic interpolation).
Key Integration Patterns:
xarray natively, pulling unit information from metadata.numpy arrays with pint units but converts them to xarray internally.xarray as the primary interface, calculations can leverage coordinate information directly from the dataset.MetPy extends xarray via an accessor that allows you to quickly parse Climate and Forecasting (CF) metadata, handle projections, and extract unit-aware arrays from multi-dimensional datasets.
Key capabilities include:
ds.metpy.parse_cf(): Parses CF metadata.ds.variable.metpy.unit_array: Returns dimensions (like lat/lon) as unit-aware arrays.ds.variable.metpy.time: Accesses time information.ds.variable.metpy.sel(...): Allows selecting data using unit-aware coordinates (e.g., selecting a specific pressure level).import xarray as xr
import metpy
from metpy.cbook import get_test_data
from metpy.units import units
ds = xr.open_dataset(get_test_data('narr_example.nc', as_file_obj = False))
ds = ds.metpy.parse_cf()
# Grab lat/lon values from file as unit arrays
lats = ds.lat.metpy.unit_array
lons = ds.lon.metpy.unit_array
# Get the valid time
vtime = ds.Temperature_isobaric.metpy.time[0]
# Get the 700-hPa heights without manually identifying the vertical coordinate
hght_700 = ds.Geopotential_height_isobaric.metpy.sel(vertical=700 * units.hPa,
time=vtime)asv (airspeed velocity) to manage performance benchmarking. It builds environments based on historical and current software iterations, runs benchmark functions, and compiles results into HTML pages. This process is integrated into the CI/CD workflow via GitHub Actions and Jenkins to identify bottlenecks and track performance changes in pull requests.