G. Roderick Singleton wrote:
On Sat, 2005-02-26 at 13:28 -0500, [EMAIL PROTECTED] wrote:You should NEVER, EVER, use a spreadsheet to do statistical computations and graphics of any kind, for any reason. Spreadsheets are not a natural way to do statistical analysis and are dangerous to use for statistical computations. You should use R: http://www.r-project.org/. Spreadsheets are dangerous to use for even storing data for input to other programs. You can check for numerous sites on the web that talk about these dangers. Most directly concern the use of Excel but many of the problems apply equally well to any spreadsheet. Spreadsheets have a certain limited usefulness but if you try to go beyond these very simple applications you are asking for problems. Use the appropriate tool for the job. If you are doing statistics and/or graphics use R.
Previously using Lotus123 I could compare a Y column of data to one or more X
columns, getting the coefficients of correlation, a constant, and a multiplier of X that produces expected value of Y. Thus Y = A + BX . A being the constant and B the multiplier of X to produce the expected value of Y for any input of X. In openoffice statistics COREL produces the coefficient of correlation,
but I have not been able to get the other values. I have tried FORECAST, which may or may not be the right one, but have mismanaged such as to get no results The same applies to LINEST under ARRAY. I must be too ignorant to solve this. Any help?
Perhaps http://www.ooomacros.org/user.php#106652 will be useful
Statistics is more than doing computations. Doing statistical analysis well requires a thoughtful analysis which R is designed, by its nature, to help provide. This is an iterative process of investigating the data and determining what the data are telling us. Plus you can trust the computations done by R to be correct.
Rick B.
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
