


VERSION 0.2.20: sub revision - ma dec  2 13:17:02 CET 2013

bugfix in the detemination of the array-position


VERSION 0.2.19: sub revision - za nov 30 22:12:39 CET 2013

adding a more advanced 'stein-alpert' example
some serious bugfixes concerning 'dsel'. Now, it is always defined from the beginning, even when it is not specified by the user ->in that case, data[i]['dsel'] it is equal to range(data[i]['dims']). 

VERSION 0.2.18: sub revision - wo nov 27 14:54:38 CET 2013

whoops, previous versions above 0.2.14 won't work


VERSION 0.2.14: sub revision - wo nov 27 10:02:56 CET 2013

add a forcearray option so that the function number of input dimensions is at least 1.


VERSION 0.2.13: sub revision - di nov 26 14:24:12 CET 2013

bugfix in the construction of the function output dimensions


VERSION 0.2.12: sub revision - zo nov 24 19:18:50 CET 2013

Add README.md to MANIFEST.in, so that it can be installed with pypi-install


VERSION 0.2.11: sub revision - zo nov 17 20:49:35 CET 2013

ignore the copyvalues option when a variable doesn't contain any values.


VERSION 0.2.10: sub revision - zo nov 17 15:06:04 CET 2013

add example for copying values
In order to support without memory problems very large NetCDF variables when copying them, the 'copyvalues' option in 'nccopyvariable' now uses pynacolada.pcd() to copy the values.
bugfix for the case when the output netcdf variable already exists


VERSION 0.2.9: sub revision - zo nov 17 09:08:39 CET 2013

clean up code to get rid of unnecessary dependencies
remove ncgetformat function. It has the same functionality as nctypecode. The latter is renamed to ncgettypecode
Some additions in the README.md
Add documentation to all the procedures
Additional example: pcdsellonlatbox.py


VERSION 0.2.8: sub revision - za nov 16 10:12:37 CET 2013

clean up type() calls
update examples: calculating the running time-average (examples/pcdfuncionexamples.py)
syntax code clean-up (pynacolada/pynacolada.py)
predim is now defined as a data input attribute

VERSION 0.2.7: sub revision - vr nov 15 12:18:37 CET 2013

load function modules from ncdfextract.py (forgotten 'return' at the end of __init.py__)


VERSION 0.2.6: sub revision - wo nov 13 15:32:40 CET 2013

bug fixes: 
- As expected by 'readicecubeps', the dimension that represents the 'list' of files should always be the first (outer) dimension!
- Create arrays with dimension-sizes corresponding to the lengths of dimnoiter-items, and not to the dimnoiter-items itself.


VERSION 0.2.4: sub revision - di nov 12 10:26:08 CET 2013

rename pcl to pcd to be more consistent to the package name and its purpose.


VERSION 0.2.3: sub revision - ma nov 11 22:19:44 CET 2013

update description of the package in README.md


VERSION 0.2.2: sub revision - ma nov 11 21:31:46 CET 2013

some readability changes in the examples.


VERSION 0.2.1: sub revision - zo nov 10 21:47:04 CET 2013

make it python3 compatible (hopefully) by providing brackets for the 'print' statements
remove unnecessary print statements
add a more extensive example with netcdf data from COSMO-CLM (examples/testana.py). This is also added to the README.md


VERSION 0.2.0: minor revision - zo nov 10 17:20:52 CET 2013

adding additional handy procedures for doing basic NetCDF operations from deprecated ncdftools package so that pynacolada becomes independent of it.
moving README.txt to README.md (github standard)


VERSION 0.1.0: minor revision - zo nov 10 15:12:39 CET 2013

Adding information on how to install package (see INSTALL)
Adding build-files to package (cfr. ./ubuntubuild.sh)
bugfix when defining the extra dimension for a file list (still needs to be tested)
adding some

VERSION 0.0.3: sub revision - vr nov  8 15:00:04 CET 2013

Add exmple directory and licence


VERSION 0.0.2: sub revision - vr nov  8 14:55:00 CET 2013


example update in README.txt

VERSION 0.0.1: sub revision - vr nov  8 14:38:12 CET 2013


Initial version. 

Current options that implemented:

 - implement dimension aliases (-> implemented as 'daliases'-option for the data input definition)
 - allow multiple input files in a stream (such as hourly climate data). Therefore, one needs to define input data as a filename-list
 - select custom coordinate frames (-> implemented as 'dsel'-option for data input definition)
 - detect output dtype, first from output data file, then from output function
 - sensible determination of memory limitation (-> maxmembytes)
 - allow wild cards in specifying function dimensions (-> appenddim!)
