[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

*Subject*: **Additional Directives to Functions to Trigger Special Behaviour**

*From*:**"Leonard Mada" <discoleo@gmx.net>***To*: office@lists.oasis-open.org, office-comment@lists.oasis-open.org*Date*: Wed, 13 Feb 2008 23:39:30 +0100

Hi all, recently I stumbled upon various inconsistencies in evaluating functions in various spreadsheet applications (especially MS vs other). Unfortunately, the interpretation of various data-types is also inconsistent within different functions in the same spreadsheet application. Especially troubling issues: - are errors propagated and - is text interpreted as a number I would not only like to see consistency within an application/s, actually I would like to be able to control/fine tune how every function behaves, because I might want the function to behave differently in different situations. I especially encourage a paradigm shift in how spreadsheets operate. There are great solutions already implemented in various programs, and I will briefly examine today one of these features. The previous problem is brilliantly solved in R (http://cran.R-project.org), where various functions can accept additional parameters (even parameters that are not used within the function but are passed further down to other functions called by the first one). e.g. one may have 'a_function'(param1, param2, '...') where '...' signifies additional parameters Also, it is possible in R to name the parameters and use them therefore in whatever order one wishes. This last option might be useful, IF some of the spreadsheet functions accept a variable number of parameters, so using '...' will be ambiguous. Therefore, I propose to extend the definition of all relevant functions to accept the following named arguments: 1.) IGNORE_ERR - ignore cells that contain an error and evaluate the function without considering that particular cell - default: IGNORE_ERR = FALSE (or IGNORE_ERR(FALSE)) - e.g MAX('various_cells_or_numebrs', IGNORE_ERR = TRUE) 2.) EVALUATE_TEXT - interpret a text as a number - default: I am not sure - e.g. SUM('various_cells_or_numbers', EVALUATE_TEXT = TRUE) NOTES ===== a.) VARIABLE(TRUE) vs VARIABLE = TRUE Because x = TRUE might be interpreted as an ordinary number (how do spreadsheets cope with this?), it might be useful to use a function (like x(TURE)) that sets the state inside the primary function. But these are rather implementation details. This additional parameter should behave more like a special object that triggers a specific behaviour of the function. b.) Spreadsheet-Level vs Function-Level I do NOT want to set these options at spreadsheet-level! Setting them at function-level one can use various evaluations within the same spreadsheet, BUT MOST IMPORTANTLY, this is portable. When a different user opens the same spreadsheet, the functions will be interpreted identically. With options set globally, this isn't possible. Sincerely, Leonard Mada -- Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten Browser-Versionen downloaden: http://www.gmx.net/de/go/browser

**References**:**Gaussian Distribution vs Normal Distribution***From:*"Leonard Mada" <discoleo@gmx.net>

**Re: Gaussian Distribution vs Normal Distribution***From:*robert_weir@us.ibm.com

**Re: [office-comment] Re: Gaussian Distribution vs Normal Distribution***From:*"David A. Wheeler" <dwheeler@dwheeler.com>

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]