v1.3.3

Aggregation of gridded data now supports partially collapsing multi-dimensional coordinates. Also fixed a bug to do with multiple variable aggregation where the masks of each variable are different (JASCIS-281).

v1.3.2

This minor release fixes an issue around plotting NASA 'blue marble' images.

v1.3.1

This minor release fixes a number of small issues, particularly around plotting.

CIS 1.3.1 updates

• JASCIS-231 & JASCIS-209 - CIS now better determines the yaxis when the user specifies the xaxis as 'time' so that overlaying multiple time series is easy
• JASCIS-283 - An issue with setting xmin or xmax using datetimes
• A minor fix to the AerosolCCI product
• Updates to support Iris 1.9.0

v1.3.0

CIS 1.3 features

• Some significant optimisations have been made in reading Caliop, CCI and Aeronet datasets, there have also been speed
  improvements for ungridded data subsetting
• New Pandas interface allows the easy creation of DataFrames through the 'as_data_frame' method on Gridded or Ungridded
  data. Pandas is an extensive python library providing many powerful data analysis algorithms and routines.
• Compatibility updates for newer versions of Numpy and SciPy. The minimum require version of SciPy is now 0.16.0
• Swapped out Basemap plotting routines for Cartopy. This removed a dependancy (as Cartopy was already required by
  Iris), and has given us more flexibility for plotting different projections in the future
• Plots now automatically try to use the most appropriate resolution background images for plots over coastlines NASA
  blue marble images.
• 'scatter_overlay' plots have been completely removed (they have been deprecated for the last two versions), the same
  functionality can be achieved through the more generic 'overlay' plots.
• Update to the UngriddedData.coord() and .coords() API to match the changes in IRIS >=1.8. This allows users to also
  search for coordinates by supplying a Coord instance to compare against. Currently this only compares
  standard names, but this may be extended in the future.

Bugs fixed

• JASCIS-279 - This release removes the basemap dependency and means we can use a much newer version of GEOS which
  doesn't clash with the SciTools version
• JASCIS-267 - Fixed ASCII file reading to be compatible with Numpy 1.9
• JASCIS-259 - Fixed Stats unit tests to reflect updates in SciPy (>0.15.0) linear regression routines for masked arrays
• JASCIS-211 - Subsetting now accepts variable names (rather than axes shorthands) more consistently, the docs have
  been updated to make the dangers of relying on axes shorthands clear and an error is now thrown if a specific subset
  coordinate is not found.
• JASCIS-275 - The ungridded subsetting is now done array-wise rather than element wise giving large performance
  improvements

v1.2.1

This update adds support for Aerosol and Cloud CCI v3 files.

v1.2.0

CIS 1.2 features

• All new cis info command provides much more detailed information about ungridded data variables and enables multiple variables to be output at a time.
• Updated a number of routines to take advantage of Iris 1.8 features. In particular gridded-gridded collocation using the nearest neighbour kernel should be significantly faster. Iris 1.8 is now the minimum version required for CIS.
• Gridded-ungridded collocation now supports collocation from cubes with hybrid height or hybrid pressure coordinates for both nearest neighbour and linear interpolation kernels.
• Built-in support for reading multiple HadGEM .pp files directly.
• All new API and plugin development documentation, including a number of tutorials

Bugs fixed

• JASCIS-253 - Any ungridded points which contain a NaN in any of its coordinate values will now be ignored by CIS
• JASCIS-250 - Multiple HadGEM files can now be read correctly through the new data plugins.
• JASCIS-197 - Gridded-gridded collocation now respects scalar coordinates
• JASCIS-199 - Aggregation now correctly uses the bounds supplied by the user, even when collapsing to length one coordinates.
• Speed improvement to the ungridded-gridded collocation using linear interpolation
• Several bug fixes for reading multiple GASSP ship files
• Renamed and restructured the collocation modules for consistency
• Many documentation spelling and formatting updates
• Many code formatting updates for PEP8 compliance

v1.2 r2

Documentation updates, moving towards 1.2 release.

v1.2 rc1

First release candidate for v1.2.

v1.1.1

This minor release primarily contained bug fixes and documentation updates.

Bugs fixed

• JASCIS-181 - Updated eval documentation
• JASCIS-239 - Documented the requirement of PyHamCrest for running tests
• JASCIS-249 - CIS will now accept variables and filenames (such as Windows paths) which include a colon as long as they are escaped with a backslash. E.g. cis plot my_var:C\:\my_file.nc.
• Occaisonly HDF will exit when reading an invalid HDF file without throwing any exceptions. To protect against this the HDF reader will now insist on an .hdf extension for any files it reads.

v1.1.0

CIS 1.1 features

• JASMIN-CIS is now called CIS, and the packages, modules and documentation have been renamed accordingly.
• Conda packages are now available to allow much easier installation of CIS, and across more platforms: Linux, OSX and Windows.
• PyHDF is now an optional dependency. This makes the installation of CIS on e.g. Windows much easier when HDF reading is not required.

Bugs fixed

• JASCIS-243 - Error when reading multiple GASSP aircraft files
• JASCIS-139 - Updated ungridded aggregation to rename any variables which clash with coordinate variables, as this breaks during the output otherwise.
• Compatibility fixes for Numpy versions >1.8 and Python-NetCDF versions >1.1.
• Fix Caliop pressure units which were stored as hPA, but need to be hPa to conform to CF.
• The integration test data has been moved completely out of the repository - making the download quicker and less bloated. It's location can be specified by setting the CIS_DATA_HOME environment variable.
• A test runner has been created to allow easy runnning of the unit and integration test.