Version 0.3 Updates¶

Version 0.3.1¶

Fixes¶

fixed issue when could not read data from an “fdb” source

Version 0.3.0¶

New features¶

added new source ads to retrieve data from the Copernicus Atmosphere Data Store (ADS). See the Retrieving data from the ADS notebook example.
added pandas and geopandas support. See the Using pandas data notebook example.
added Caching policies. See the Cache policies and Configuration notebook examples.
changed the return type of data(), which now returns an ndarray. Previously it returned a tuple of ndarrays. See the GRIB: getting latitudes, longitudes and values (regular LL grid) notebook example.
added data() method to FieldList. See the GRIB: getting latitudes, longitudes and values (regular LL grid) notebook example.
added the valid_datetime metadata key, which can be used in Field.metadata() and FieldList methods like metadata(), sel() and order_by() etc. It is particularly useful for GRIB data because the this piece of information was previously only available as two separate keys (validityDate and validityTime).
```
>>> import earthkit.data
>>> ds = earthkit.data.from_source("file", "docs/how-tos/test.grib")
>>> ds[0].metadata("validityDate")
20200513
>>> ds[0].metadata("validityTime")
1200
>>> ds[0].metadata("valid_datetime")
datetime.datetime(2020, 5, 13, 12, 0)
```
implemented FieldList for NetCDF data. See the NetCDF: working with fieldlists example.
added experimental Metadata handling. See the /how-tos/legacy/metadata.ipynb example.
added experimental input transformer tools for earthkit subpackages. It uses type-setting, or explicit mapping, to try to ensure that the inputs passed to a function are converted to the appropriate type for that method. This means that earthkit users do not have to worry about the format of their data.

Fixes¶

fixed issue when cache initialisation hanged when log level was set to debug.