[R] CRAN and Multiple Linear Regression

Anja und Th. Sponsel anjaspp at netscape.net
Wed May 14 00:54:50 CEST 2008


I just have read the guide and I can do some small steps with cran but
I still have no clue...

I have data like this:
	X1	X2	X3	...	X21	Y
1	0	0	0	...	18	-0,07254
2	1	0	0	...	6	-0,14921
3	0	1	0	...	12	-0,04165
4	0	0	0	...	8	0,08359
5	...	...	...	...	...	...
120	0	0	1	1	8	0,07928

My script is:
> require(stats); require(graphics)
> test = read.table("test.dat", header=T)
> x <- test[, 1:21]
> y <- test[, "Y"]
> lm(y ~ x)

The error may be an invalid type (list) for variable x (see below):
Fehler in model.frame.default(formula = y ~ x, drop.unused.levels =
TRUE) :
  ungültiger Typ (list) für die Variable 'x'

Then I tried:
...
> lm(data=test)
Then I get a lot of coefficients, but I'm not sure. That cant be the
result.

Then I tried:
> summary(lm(data = wetter))

Call:
lm(data = wetter)

Residuals:
ALL 120 residuals are 0: no residual degrees of freedom!

Coefficients: (139 not defined because of singularities)
                       Estimate Std. Error t value Pr(>|t|)
(Intercept)            0.044747         NA      NA       NA
X2                     0.660275         NA      NA       NA
X3                    -0.097629         NA      NA       NA
X4                     0.647851         NA      NA       NA
X5                    -3.305631         NA      NA       NA
...                    0.249788         NA      NA       NA

I am confused...

Best regards,
 Thomas



More information about the R-help mailing list