ChangeLog

CHANGES
=======

0.7.3
-----

* dereference pandas array from pint pandas

0.7.2
-----

* increase min version on excel handler

0.7.1
-----

* fix: series initialisation from dimensionless variable
* big refactoring - moved framework files from teapot to ngmodel
* additional bugfixes after refactoring to pint-pandas

0.7.0
-----

* use pint-pandas instead of Quantities
* new csv importer for Ngmodel server
* fixed average chart logging improvements
* added parameter name to 'zero values' warning
* two new charts: input vars and grid

0.6.8
-----

* bug fix: array of "ref value"s must be sorted by datetime
* increment excel helper version

0.6.7
-----

* variability of uncertainty function is now relative to initial value of mean
* started added documentation on excel format v2 bug fix
* increased version number of excel helper to 0.6.6

0.6.6
-----

* added openpyxl as extra
* removed old module moved openpyxl dependency into method

0.6.5
-----

* fixed pbr package declaration

0.6.4
-----

* cleaned up imports

0.6.3
-----

* cleaned up authors
* moved to pbr for build

0.6.0
-----

* use extrapolation with scipy interp
* new interpolation function
* minimum working version
* increment version number
* with scenarios
* with idle energy consumption
* make graph file type a cmd line setting
* added a mention of CAGR to the readme
* truncate ref date
* removed debug print statesmtenst
* truncate ref date to 1st of month
* print error message when non-string date
* - new
* - inc version number
* - extend time bounds to include ref date for sample generation
* - upgrade version to 0.5.7
* - fix broken ref date parsing
* - refactoring: store simulation results in subfolder output/model name / datetime - server: new page compare\_runs/model name to compare model runs
* - helper 0.5.6: new function ParameterRepository.clear\_cache
* changes to param repo exists method to accomodate scenario parameters bump version to 0.5.5
* fix random function initialisation: ignore None and empty strings
* fix random function initialisation: ignore None and empty strings
* Fixed error in generation of random distributions with zero param values
* up version number
* put xlsx2csv back in
* removed xlsx2csv
* removed xlsx2csv
* remove test of start/end date
* first tests completing
* updated for XLWings
* added XLWings ExcelHandler
* fixed choice mean generation
* increased version number
* added sphinx doc
* added option to sample\_mean\_value from distributions
* added option to sample\_mean\_value from distributions
* added option to sample\_mean\_value from distributions
* Added documentation to ExcelParameterLoader
* more tests
* fixed excel param distribution generation
* populate empty fields from default
* with working implementation of excel loader and multiple scenarios and tags
* with working implementation of single dimension and time series distributions
* tests for distribution generation, and adding and getting parameters
* fixed test cases
* fixed value generation when size is None version inc to 0.3.2
* with complete serialisation and deserialisation
* proper test case for 'single\_var' kwarg add xarray to dependencies inc to v 0.3.0
* proper test case for 'single\_var' kwarg add xarray to dependencies inc to v 0.3.0
* proper test case for 'single\_var' kwarg add xarray to dependencies inc to v 0.3.0
* Added new flag 'single\_var' to freeze all variables except one to their mean value - use in sensitivity analysis
* added test fixed install dependencies
* added logging
* added logging
* added documentation
* added python 2.7 support back in
* added excel data source reload now python 3 only
* catch non-initialised item access to prepare item for access

0.2.0
-----

* added documentation and python 3 classifier
* changed README to md
* added documentation
* make python 3 ready
* import on new machine from FS

v0.1.9
------

* added CAGR support added from df support
* added CAGR support added from df support
* Initial commit