v1.5.0

• e3sm_diags can now be ran as a container, in addition to being manually installed and ran that way.
  • View how to do so on Edison at NERSC via Shifter here.
  • The auxiliary script to run a container,e3sm_diags_container.py, supports Docker and Shifter.
  • Singularity support will be added in the next release or so.
  • A quick guide will be added for Docker and Singularity soon as well.
• Scripts added for creating the time-series observational data.
• Minor additions to the viewer.

v1.4.0

• Added selectors parameter.
  • View the documentation here.
• Major bug-fix regarding issue #156, where too many threads were used.
• Added scripts to create climo and timeseries data for the observational data. Located in the analysis_data_preprocess folder.

v.1.3.4

• Fixed major issue where NetCDF-4 files would not open on Edison compute nodes.
• Added output_format_subplot parameter to save individual panels when using matplotlib for the plotsets:
  • lat_lon
  • polar
  • zonal_mean_xy
  • zonal_mean_2d
  • cosp_histogram
• Expanded provenance. Now for each run, the command used, the py file, and the cfg file used are save in {results_dir}/prov. A yml file of the environment is also saved there as well.

v1.3.3

• Finally renamed package to e3sm_diags.
  • But acme_diags is still supported for those who update and don't read the changelogs.
• Removed mesalib as a dependency.
• Other very minor fixes.

v1.3.2

• Python 2 and 3 support.

v1.3.1

• CDAT 8 support.
• Use conservative regridding by default.
• Expanded provenance:
  • Dump of which dependencies are used, stored in results_dir/env_prov.txt.
  • All parameters work via the command line.
  • --no_viewer parameter to not create the viewer.

v1.3.0

• Stronger support for model vs model run with a new parameter run_type
• Cleaner code directory with all default configuration files converted to .cfg
• Add AOD550, surface wind stress and surface air temperature as default variables
• Several bug fixes

v1.2.1

• The provenance works better. The command should only create one plot.
• Running with the mpl backend on a headless machine should work.

v1.2.0

• Added Taylor Diagrams for spatial variability for annual and seasonal means.
• Added provenance: a single run can be recreated with a single command.
• A dependency, cdp, was updated to support the provenance feature.
  • If using a development environment, explained here, create a new environment by following the instructions again.
  • If using a regular environment, just run conda update acme_diags to get v1.2.0 of acme_diags.

v1.1.1

• Implemented a workaround for multiprocess running with VCS as the backend