|Jul 17, 2017|
|3 years 6 months ||Making the FitMult class as similar as possible to FitModel
|3 years 6 months ||The new Trend class now has a toString() method
|3 years 6 months ||FitModel::min_nb_pts is not const anymore since this was required
for properly handling user-defined lapse rates.
There were possibilities of using nodata in the computation of
ILWREPS, this is not properly tested for. The documentation has been
|Jul 14, 2017|
|3 years 6 months ||The multiple linear interpolation class has a new structure, more
along the lines of the other statistical regression models. A few
more checks have been implemented. Tests conducted for precipitation
in the Valais are very promising: the R2 jumps fr
|Jul 12, 2017|
|3 years 6 months ||More documentation, more checks for the multiple linear regression
|Jul 11, 2017|
|3 years 6 months ||The R2 is now computed in another way, the Beta0 was forgotten and
is now computed and more checks are performed on the input data.
|Jul 10, 2017|
|3 years 6 months ||A new class has been implemented that performs multiple linear
regressions. The syntax of getParams() has been changed.
|Jul 7, 2017|
|3 years 6 months ||Oups, the debug code was still around...
|3 years 6 months ||Better error messages and fixed a possibly uninitialized variable
|Jul 6, 2017|
|3 years 6 months ||Better support for the grids: fixed units conversions are assumed
and the ISWR is reconstructed from dir+diff
|3 years 6 months ||Code cleanup, constifying some variables
|3 years 6 months ||The Oshdio plugin has been cleaned up and can now read gridded data
|Jul 5, 2017|
|3 years 6 months ||The exception messages should now be much better on Windows
|3 years 6 months ||The SUPPR filter now allows supressing dates within given date
ranges, one a per station basis.
|Jul 4, 2017|
|3 years 6 months ||Cleanup before bigger changes
|3 years 6 months ||For the ADD and the MULT filters, it is now possible to use
multi-columns corrections files and to specify which column should
|Jul 3, 2017|
|3 years 6 months ||Paving the way for multivariate regressions: the whole vecMeta is
passed to the Trend class
|Jun 30, 2017|
|3 years 6 months ||Documentation improvements
|3 years 6 months ||The handling of detrending / retrending is now much cleaner, only
one class is responsible for it.
|Jun 29, 2017|
|3 years 6 months ||Simplifying the prototypes of the spatial interpolation algorithms.
When there is no need of a GridsManager or a Meteo2DInterpolator (as
is most often the case), these are not provided to the constructor.