Py-ART: Python ARM Radar Toolkit

2.2.0 · active · verified Thu Apr 16

Py-ART is an open-source Python library designed for working with weather radar data. It provides tools for reading, processing, visualizing, and analyzing data from a variety of radar formats, including NEXRAD, UF, and ODIM_H5. Version 2.2.0, released recently, continues to build on its foundation with improvements in data handling and integration with other scientific libraries like xradar and cmweather. The project maintains an active release cadence with minor releases and bug fixes.

Common errors

```
ImportError: No module named 'pyart'
```
cause The Py-ART library is not installed in your active Python environment.

fix Run `pip install arm-pyart` to install the library.
```
ModuleNotFoundError: No module named 'cartopy'
```
cause A core dependency of Py-ART, `cartopy` (or `netCDF4`, `xarray`, etc.), is missing or failed to install correctly due to underlying system dependencies.

fix Install the missing package directly (e.g., `pip install cartopy`). If `pip` fails due to binary issues, consider using `conda install -c conda-forge pyart` which often resolves complex dependency chains by providing pre-built binaries.
```
AttributeError: 'Radar' object has no attribute 'gate_altitude'
```
cause Changes in the internal data model or how data attributes are exposed, particularly after the `xradar` integration in v2.0.0, may mean previously accessible attributes have moved or been renamed.

fix Consult the Py-ART v2.0.0 release notes and updated documentation regarding changes to the `Radar` object's attributes and methods. Use helper functions or updated methods if available, or access data through the `xarray` dataset wrapped by the `Radar` object.
```
TypeError: plot() got an unexpected keyword argument 'cmap'
```
cause Py-ART transitioned to using `cmweather` for colormaps in v2.0.0, which may alter how colormaps are passed to plotting functions or require specific `cmweather` colormap objects.

fix Verify how colormaps are now handled with `cmweather`. You might need to import colormaps from `cmweather.colormaps` or ensure your custom colormaps are compatible with the new plotting interface.

Warnings

breaking Py-ART v2.0.0 introduced significant internal changes, primarily adopting `xradar` as the main backend for reading radar datasets and utilizing `cmweather` for colormaps. Code relying on direct internal reader implementations or Py-ART's older colormap handling may break.
Fix: Review data loading routines, especially for new formats, and adapt to `xradar`'s approach. For colormaps, import from `cmweather` (e.g., `cmweather.colormaps`) or verify existing colormaps are still available and behave as expected.
breaking As of v1.19.3, Py-ART officially requires Python 3.10 or newer. Installations on older Python versions will fail or result in `ImportError`.
Fix: Upgrade your Python environment to version 3.10 or newer. It is recommended to use `conda` or a virtual environment for managing Python versions.
gotcha Installation of geospatial dependencies like `cartopy` and `netCDF4` can be challenging, often requiring system-level libraries (e.g., GEOS, PROJ, HDF5). `pip` installations might fail without these prerequisites.
Fix: For `cartopy` or `netCDF4` installation issues, consult their respective documentation for system-level dependencies. Using `conda` (`conda install -c conda-forge pyart`) is strongly recommended, as it often resolves complex binary dependency chains automatically.

Install

pip install arm-pyart Default Install

Imports

Radar
wrong:
```
from pyart.core import Radar
```
correct:
```
import pyart; radar = pyart.io.read('file.nc')
```
While `pyart.core.Radar` is the underlying class, `pyart.io.read` is the standard factory function for loading radar data into a Radar object from a file, which is the most common use case.
RadarDisplay
wrong:
```
from pyart.graph.radardisplay import RadarDisplay
```
correct:
```
import pyart; display = pyart.graph.RadarDisplay(radar)
```
Prefer accessing `RadarDisplay` via the `pyart.graph` module (e.g., `pyart.graph.RadarDisplay`) for better readability and to follow typical Py-ART usage patterns.
read
```
import pyart; radar = pyart.io.read('file.nc')
```
The primary function for loading various radar file formats into a Radar object.

Quickstart

This quickstart demonstrates how to create a dummy radar object and plot a basic Plan Position Indicator (PPI) reflectivity slice using Py-ART's `RadarDisplay`. This allows for immediate visualization without needing external data files.

import pyart
import matplotlib.pyplot as plt

# Create a dummy radar object for demonstration
radar = pyart.testing.make_empty_ppi_radar(10, 10, 10)

# Initialize a RadarDisplay object
display = pyart.graph.RadarDisplay(radar)

# Plot a PPI (Plan Position Indicator) slice of reflectivity
fig = plt.figure(figsize=(8, 6))
ax = fig.add_subplot(111)
display.plot('reflectivity', 0, title='Simulated Reflectivity (PPI)', ax=ax)
plt.show()

view raw JSON →