Version 1.0.0

Released on ECMWF desktops, ecgate and lxc on 2018-12-21. Also on github and pypi (pip install metview)

This version is the first 'full release', containing only some internal cleanups. Some more example Jupyter notebooks were added.

From this point, all developments and issue tracking will be performed directly on github. Please report any issues there.

Version 0.9.1

Released on ECMWF desktops, ecgate and lxc on 2018-11-24. Also on github and pypi (pip install metview).

This version is considered beta.

Bug

• [MPY-207] - Filedset slicing fails with a large number of fields

Task

• [MPY-205] - Allow the creation of an empty Fieldset object
• [MPY-206] - Add append() function to Fieldset class
• [MPY-208] - Passed named arguments to dataset_to_fieldset function
• [MPY-209] - Ignore errors on exit from a data examiner
• [MPY-210] - Add expanded field-obs example

Version 0.8.6

Released on ECMWF desktops, ecgate and lxc on 2018-10-26. Also on github and pypi (pip install metview)

Bug

• [MPY-186] - Cannot plot a multi-page PostScript file - now we can do it by inserting calls to newpage():
  • e.g. plot(dw, data1, visdef1, newpage(), data2, visdef2) - requires Metview >=5.2
• [MPY-196] - Cannot take scalar numpy numbers (e.g. np.int64) as input arguments to Metview functions

Task

• [MPY-136] - Use environment variable to determine how to invoke Metview (METVIEW_PYTHON_START_CMD)
• [MPY-137] - Support passing of float32 vectors between Metview and Python numPy
• [MPY-153] - Add function to convert NetCDF to xarray dataset
  • e.g. xs = mv.mcross_sect(line=[59,-180,-13,158],data=grib); xarr = xs.to_dataset()
• [MPY-162] - Tidy up value_from_metview
• [MPY-163] - Tidy up push_arg
• [MPY-177] - Revise example scripts in repo
• [MPY-187] - Allow assignment to indexed fields in a Fieldset object
  • e.g. fs[3] = fs[2] + 5
• [MPY-188] - Allow xarray datasets to the provided as input to Metview functions
  • e.g. m = mv.mean(my_xarray_dataset); also f = mv.dataset_to_fieldset(x) - only for very compliant datasets though!
• [MPY-189] - Check and revise the scripts in the examples folder
• [MPY-190] - Create an example using data from the CDS
• [MPY-191] - Ensure that FLEXTRA work from Metview's Python API
• [MPY-192] - Create Jupyter notebook examples
• [MPY-193] - Fix netcdf and geopoints functions that take indexes as parameters
• [MPY-194] - Implement self-check command to ensure Metview installation ok
  • python -m metview selfcheck
• [MPY-195] - Ensure all index-returning functions are context-sensitive
• [MPY-197] - Add ability to export table data as a Pandas Dataframe
  • df = table.to_dataframe()
• [MPY-198] - Add batch plotting tests
• [MPY-199] - Improve organisation of code in bindings.py
• [MPY-200] - Remove experimental 'mf' class
• [MPY-201] - Support extended slicing on Fieldset objects 
  • e.g. fields = fs[2:12:3]

...

Released on ECMWF desktops, ecgate and lxc on 2018-08-01. Also on github.

Bug

• [MPY-164] - importing the metview module overrides the keyboard interrupt signal
• [MPY-179] - Cannot find libMvMacro.so in /usr/lib

Task

• [MPY-180] - Replace klass with class_ in MARS requests (change in behaviour)
• [MPY-183] - Use cfgrib in metview-python

...

Released on ECMWF desktops and ecgate on 2018-06-08.

Task

• [MPY-144] - Support passing of geopointset between Macro and Python
• [MPY-150] - Add ability to export ODB table as a Pandas Dataframe
• [MPY-151] - Simplify data classes through multiple inheritance
• [MPY-152] - Allow access to individual geopoints
• [MPY-154] - Implement aliases for 'class' as MARS parameter
• [MPY-155] - Improve error message if correct version of Metview not found

...