regular update for matplotlib 3.9
- add tests for mpl 3.9, see !43
- the colorbar tests as updating the colorbar is currently broken with matplotlib 3.9, see #35
Compatibility fixes and code formatting
- fixed compatiblity with latest matplotlib version, see !38
- fix image comparison and add reference and test images as CI artifacts, see !37
Compatibility fixes
- We now use the new
convert_coordinate
method that has been introduced with psyplot v1.4.1 (see psyplot/psyplot#39 and #30)
- psy-simple is now compatible with matplotlib 3.5 (see #31)
Compatibility fixes and LGPL license
As with psyplot 1.4.0, psy-simple is now continuously tested and deployed with CircleCI.
- Compatibility fixes for matplotlib>=3.3
- psyplot does now have a CITATION.cff file, see https://citation-file-format.github.io
- psy-simple is now officially licensed under LGPL-3.0-only, see #28
- We use CicleCI now for a standardized CI/CD pipeline to build and test the code and docs all at one place, see #27
Patch for plotting the polygons with 3D bounds
- the values
'tri'
,'tricontour'
and'tricontourf'
for theplot
formatoptions have been depreceated and should be replaced by'poly'
,'contour'
and'contourf'
respectively, see #23
- A bug was fixed with the
extend
formatoption ifplot=None
, see #20 - variables with 3D bounds are now interpreted correctly, see #24
New background and mask formatoptions and more options for colorbar bounds
a new
background
formatoption has been implemented that allows to set the facecolor of the axes (i.e. the background color for the plot)a new
mask
formatoption has been implemented that allows to mask the data based on a mask that can either be in the dataset or in a separate file (see #15)the
bounds
and otherticks
(e.g.xticks, yticks, cticks
) formatoptions have gained multiple new values (all backwards-compatible, see #13):they now support discrete logarithmic and symmetric bounds/ticks via
bounds='log'
andbounds='symlog'
.The bounds and other tick formatoptions (xticks, cticks, yticks, etc.) now support a dictionary as a value, e.g.:
plotter.update(bounds={'method': 'rounded', 'percmin': 5})
You can specify
vmin
andvmax
for color bounds and ticks which prevents their automatic estimation, e.g. via:plotter.update(bounds={'method': 'rounded', 'vmin': 50, 'vmax': 75}) # or plotter.update(bounds=['rounded', None, None, None, 50, 75])
- values in the statusbar are only shown, if the drawn artist contains the cursor position, see #18
- psy-simple now requires matplotlib greater or equal than 2.0
- psy-simple has been moved from https://github.com/Chilipp/psy-simple to https://github.com/psyplot/psy-simple, see #7
- The color handling of the color formatoption has been changed to allow appending of new data. The colors attribute can be extended by the color_cycle using the extended_colors attribute (see #10)
- Fixed a bug to calculate color bounds for uniform data, see #9
- An issue has been fixed with the setting of colorbar ticks after updating the colorbar bounds (see #13)
- The
plot2d
plotmethod now also supports unstructured data of any shape (see issue#6) - Added a
categorical
formatoption to thebarplot
plot method to allow a switch between categorical and non-categorical plots - The lineplot method now also support
'stacked'
plots
- Changelog
interp_bounds
formatoption for theplot2d
plot method (see the docs)- Added the
fldmean
plot method that can be used to directly calculate and plot the mean over the x- and y-dimensions
- The xlim and ylim formatoptions now consider inverted x- and y-axes