No regression line formulas in Office.org calc
MSchwartz at mn.rr.com
Fri Jun 10 16:09:57 UTC 2005
Major Seery wrote:
> I spent a couple hours yesterday fiddling with R, and you're right, the
> command line environment is more like a brick wall than a mere hurdle. I
> check out the GUI's you and others suggested.
Keep in mind that R is a full fledged programming language. While the
GUIs attempt to shield users from that aspect of R, they are also
"confining". It would be the same as trying to code in C or FORTRAN
using only a menu based approach to build an application.
"One must learn by doing the thing. For though you think you know it,
you have no certainty until you try" - Sophocles
If you want to stay with R and use a GUI based approach, I do honestly
think that John Fox' R Commander is probably the easiest in terms of a
"point and click" menu based mechanism. Being based on tcl/tk, it is
also importantly cross-platform.
If you want a good starting point using R from the command line, you
won't find a more gentle introduction to it than Peter Dalgaard's
Introductory Statistics With R:
Another good reference is An Introduction To R, which is available at:
Finally, there is also excellent Contributed Documentation at:
Lastly, if all else fails and you wish to revert to a spreadsheet, I
would suggest considering Gnumeric. They have gone to some lengths to
react to criticism of the math/stat functions that they have included:
They also tend to stay within the confines of known standards, such as
the IEEE 754 floating point representation standard (which the OO.org
folks have not, in their zeal to replicate Excel's functionality in this
The Gnumeric developers have also consulted at times with members of the
R Core team on various issues.
I do not advocate the use of spreadsheets for statistical work, given
the known and well publicized limitations:
More information about the users