Commit commit:0e2352a
Contents, since 1.2.0:
Functionalities
Extract a subzone, keeping projection properties: field.extract_subarray()
Update of geometry.plotgeometry()
Factorization of functions in domain_maker
Histograms of fields: field.hist() + epy_hist.py
Option epy_stats.py --only_maxdiff/--oxd
Horizontaly averaged profiles (FA, epy_profile.py)
Grid-to-grid resampling: field.resample() and .resample_on_regularll()
FA
Ability to read/write GRIB2-encoded fields in FA (using grib_api), with an arpifs4py.so from cy42_op1 onwards
netCDF
On-the-fly options in writing: writefield(adhoc_behaviour=...)
Flattening
EPYWEB
Mode filesystem
Numerous various fixes
usevortex
Points/profiles extraction to netCDF: usevortex.extractor()
Technical
Unit-tests
Preparation to future compatibility with Python3
Support for LFI on MF-Bull
Miscellaneous
Numerous various fixes
Field data accessors, made cleaner
Various other cleanings
Update of the documentation
Notebooks in Tutorial
Commit commit:e0dcee6
netCDF
Several improvements w.r.t. convention CF-1.6 and geometry.
(Almost full-) Compatibility with the forthcoming SURFEX v8.1 netCDF outputs.
EPYWEB
Integration of EPYWEB: a Vortex+Epygram HTML interface for plotting fields from MF operational suites and OLIVE experiments.
Other new features
Pre-staging mode in usevortex
Function fields.psikhi2uv()
Function util.datetimerange()
Computation of wind direction
Miscellaneous
Porting fixes for Mageia6
More flexible colormaps scalings (as radar)
And a quantity of fixes !...
Commit commit:b53faebc
Advanced netCDF support:
Read/write from/to netCDF files well better handled (examples to come).
Support for any kind of fields, including temporal dimension (from a single Point to a 4D field).
Facility: any field can be dumped to a netCDF file, thanks to the dump_to_nc() method.
Fields' internal data is 4D
Although rather transparent to the user, the internal data storage of fields is now always 4D. This modification brings great simplification in many places.
The getdata()/setdata() methods are hereby essential to use, with the optional d4=True.
Better handling of time dimension in fields
Corrections, modifications linked to previous item, and possibility to extract a time index or extend time dimension with another field (cf. methods).
Animation of fields
Fields with a temporal dimension now propose a plotanimation() method, that create a matplotlib animation, that can be saved to a MP4 file.
"Special resources": or how to get 3D/4D fields from 2D resource(s)
A new proxy function to create easily "special resources" (epygram.resources.special_resource()), that enable to:
- make 3D fields from a resource containing a series of horizontal fields distributed on the vertical;
- make 2D fields + temporal dimension from a series of 1-time resources;
- combination of both.
Better handling of matplotlib's figures and axes in plotfield()
Enables to have multiple plots on the same figure, and a more trustful superposition of plots.
Derivatives for Gauss grid fields (arpifs4py)
Similarly to the cartesian grids, through the compute_xy_spderivatives() method.