Gottfried Helms schrieb:
...
> > Then I tried it with a classic data set from Consumer Reports listed at
> >
> > http://lib.stat.cmu.edu/DASL/Datafiles/Cars.html
> >
> > > corr.reg(Weight,MPG) [1] 0.4684389
> > > corr.reg(Drv.Rat,MPG) [1] 0.2145594
> > > corr.reg(HP,MPG) [1] 0.1312272
> > > corr.reg(Displ,MPG) [1] 0.5197843
> > > corr.reg(Cyl,MPG) [1] 0.2893959
> >
> > I think I have the algorithm coded properly, but Dr. Chambers or anyone else could
>double check these results very easily.
> >
>
> Hmmm. Don't know why, I get completely different results. Compare that values with
>the
> first row of the table of bivariate tests below:
Upps- I had an error transferring the html-page into data...(only 3 values from MPG
cutting first digit - something said about robustness?)
With the corrected input I get the same results as Steve.
MPG Weight Drive_Ratio Horsepower
Displacement Cylinders
MPG 0,00 0,47 0,21 0,13
0,52 0,29
Weight -0,47 0,00 -0,17 -0,04
0,09 -0,08
DriveRatio -0,21 0,17 0,00 0,12
0,17 0,16
Horsepower -0,13 0,04 -0,12 0,00
0,32 -0,21
Displacement -0,52 -0,09 -0,17 -0,32
0,00 -0,29
Cylinders -0,29 0,08 -0,16 0,21
0,29 0,00
Negative values indicate that the column-variable causes the row-variable
Gottfried.
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
