New FORECAST.ETS function in Calc

[dev-list LO]

Hi Eike,

As the making of the new FORECAST.ETS functions is nearing its 
completion, I would like to discuss a naming issue.
Exponential Triple Smoothing has two types of calculation, the 
difference being the way in which the periodical (seasonal) influence is 
defined.
The first (and oldest) way is to define the periodical influence in 
absolute values (i.e. in July 10 extra ice creams are sold as compared 
to the average). This is called the additive method and is used by Excel 
2016 in their FORECAST.ETS functions.
The second (newer and often more realistic) way is define the periodical 
influence in relative terms (i.e. in July 5% extra is sold). This is 
called the multiplicative method.
The method names indicate the way forecasts are calculated.
A sample document with the functions for various data sets and graphs is 
attached to tdf#94635: 
https://bugs.documentfoundation.org/attachment.cgi?id=121750

I have created both additive and multiplicative FORECAST.ETS (forecast), 
FORECAST.ETS.STAT (statistical values like standard deviation) and 
FORECAST.ETS.PI (prediction intervals, an indication of the accuracy of 
the forecast) functions. That are 6 functions, with long names, e.g. 
FORECAST.ETS.STAT.MULT().
As an alternative I could add an optional argument to the function, 
indicating whether (1) additive, (2) multiplicative or (3) the best of 
the two is to be used for forecasts. IMHO the default value of the 
argument should be 3.
On export/import to/from xlsx, this extra argument can be 
removed/inserted with value 1.

Using this extra argument reduces the number of functions names, the 
length of the function names and quite some lines of UI code.

Before making any changes, I would like to hear your thoughts about 
this, as well as thoughts from others from a UI point of view.

Winfried

BTW, I think these functions should be proposed to be added to the 
OpenFormula standard, they have a lot of practical use cases.