Unit test updates

This micro release increases the total unit test coverage to 87.6% and fixes some bugs when doing solar calculations at high latitudes.

Removing cartopy as dependency

Cartopy was intended to be an optional dependency but it was listed as a dependency in the requirements.txt file. Removing that dependency so that other packages using ACT doing have to install.

Bug fixes

A few bugs are fixed in this micro release

• Fixed bug when trying to plot multiple skew-t plots
• Fixed bug in QC clean script
• Fixed bug with keeping the skyfield ephem file open when multi-processing is used

Additions

• Method to convert units in an object
• Adding features to make written netcdf more cf complient

ACT Version 0.6.0

This release removes astral as a dependency in favor of skyfield which greatly reduces the complexity of having different versions of astral on pypi and conda-forge. Many of the other additions were focused around improving the unit tests for ACT, especially on the visualization testing. Also removed the dependency on Boto3 and the AWS bucket for testing. Finally, improved the contour plotting module to allow for plotting of different levels such as with soil moisture. This opens up the possibility for easier 3D-contour plotting.

Bug fix for conda release

Updating to make sure both versions of astral can be used.

Basic fixes, updates to plotting, and new functions

There were a number of fixes and improvements implemented in this release

• Bug fix for when cartopy is installed but not importing correctly
• Bug fix in the cog/sog calculations when "lon" variable is not available
• Added additional error handling
• Additional unit tests for the utilities modules, increasing code coverage by 1.1%
• Improvements for the Py-ART utility to handle RHI data
• Update to MPL depolarization ration calculation
• Improved handling of ARM DQR quality control

New Features include

• Ability to specify a specific time for downloading ARM data
• New functions to perform a variety of irradiance calculations
• Adding the ability for the users to plot up the FFT shading test results

Version 0.4.3

Bug fix for conda release and requiring matplotlib 3.2.2 or less. Matplotlib 3.3.1 is problematic at the moment.

Version 0.4.2

Update to use multiple versions of Astral to fix the conda install problems.

Version 0.4.1

Updates to environment file to work with conda on new astral version.

Version 0.4.0

There were a number of feature additions and bug fixes in this version.

New Features

• Ability to plot QC for 2D variables containing 1D data (soil moisture, etc)
• Autogenerating DOI through Zenodo
• Generic writer to clean up dataset before writing to file
• Coveralls to check code coverage
• Binary reader for micropulse lidar using mpl2nc which is now an optional dependency
• Utility to convert ACT object to PyART radar object for plotting of PPI/RHI data (mpl, doppler lidar)
• Utility to calculate day/night and added examples for calculating day/night averages
• Ability to better set the y-range on timeseries plots (y_rng)

General Updates

• Updated examples and documentation
• Updated to use astral 2.2

Bug Fixes

• Fix to geoplot to add gridlines to some projections
• Revamped configuration files to deal with MetPy and Cartopy issues
• Fixed examples when MetPy or cartopy are not available
• Bug fix for the wind rose plotting

Thank you to @kenkehoe @zssherman @cgodine for the contributions!