[Date Index] [Thread Index] [Author Index]

Q. re results from Regression[ ]

• To: mathgroup at smc.vnet.net
• Subject: [mg8518] Q. re results from Regression[ ]
• From: elsner at avalon.msfc.nasa.gov (Ron Elsner)
• Date: Thu, 4 Sep 1997 02:20:13 -0400
• Sender: owner-wri-mathgroup at wolfram.com

The Statistics`Regression` has a nice function called Regression 
that can be used to do weighted linear least squares fits providing 
lots of statistical information about the results of the fit.  The 
option ParameterTable provides "standard errors" for the best fit 
parameters.  However, the package description does not define 
standard errors.  I am not familiar with this term.  Even if it 
means standared deviation, what is its mathmetical definition in 
this package?

--Ron Elsner
NASA/MSFC