Release notes

Release notes#

v0.2.0 (unreleased)#

Changes#

  • tams.data.download_examples() has been removed (PR69). Example data files now download automatically when you use tams.data.open_example() or tams.data.load_example(), and to the user cache directory instead of into the package. These functions provide a unified interface to access the example datasets, as opposed to the separate load_example_* functions previously available. This functionality requires pooch, in addition to gdown.

  • tams.load_mpas_precip() has been removed (PR70). This function name was confusing because “precip” here was a reference to the PRECIP field campaign (summer 2022). It was specifically designed to load postprocessed outputs from near-real-time runs associated with that campaign.

  • tams.contours() has been renamed to tams.contour() and now returns a GeoDataFrame of contour lines (instead of a list of arrays of line segment coordinates; PR74).

  • tams.identify() now returns a single of list of GeoDataFrames instead of also returning the identified cores (PR84). In these frames, the core column contains the cold cores within each CE (renamed from cs219), with area area_core_km2 (renamed from area219_km2). “235” and “219” language was generally removed in favor of the more general “CE” and “core” terminology, since different thresholds can be used. Scalar geometries of the core column are now either MultiPolygon, Polygon, or None (no cores) instead of always being MultiPolygon.

  • In tams.identify(), the size_filter Boolean parameter is deprecated (use size_threshold=0 instead to disable size filtering; PR84).

  • tams.calc_ellipse_eccen() has been renamed to tams.eccentricity() (PR85). The old function name can still be used, but it is deprecated and warns.

  • The minimum GeoPandas version is now 1.0.0 (PR87).

New features#

  • The existing example datasets were updated to improve compression, and additional datasets have been added, including the MOSA test cases and a sample IMERG dataset (PR70). See Example datasets for details.

  • Contouring (tams.contour()) now drops non-closed contours by default and computes whether the contour encloses higher or lower values (to account for holes in CE polygons; PR74).

  • Convex hulling can be disabled in tams.identify() (PR74).

  • Logger level and handler can be controlled using tams.set_options() and should still work even when using parallel=True. The logs now include more information about the contouring, including reasons for exclusion (PR74).

  • Create idealized datasets for testing using tams.idealized (PR80).

  • tams.fit_ellipse() can be used to obtain the ellipse fit parameters used for the eccentricity calculation (PR85). It returns a named tuple (Ellipse).

  • Specify cmap in tams.plot_tracked() (PR89).