Hi
r-help-bounces at r-project.org napsal dne 22.10.2007 14:01:14:
> Hi,
>
> I need to calculate either Error or Normalized values based on the
following> principle:
>
> Error = Observed value - reference value
> Normalized value = Observed Value - Part average
>
> appraiser <- rep(rep(1:3,c(3,3,3)),10)
> trail <- rep(rep(1:3,3),10)
> part <- rep(1:10,c(9,9,9,9,9, 9,9,9,9,9))
> value <- c(rnorm(9, 57.01,0.01), rnorm(9, 57.06,0.02), rnorm(9,
57.10,0.04),> rnorm(9, 57.07,0.03), rnorm(9, 57.12,0.025),
> rnorm(9, 57.02,0.011), rnorm(9, 57.03,0.02), rnorm(9,
57.08,> 0.013), rnorm(9, 57.01,0.06), rnorm(9, 57.03,0.015))
> off <- cbind(appraiser, trail, part, value)
> off <- data.frame(off)
> off$appraiser <- factor(off$appraiser)
> off$trail <- factor(off$trail)
> off$part <- factor(off$part)
>
> par(mfrow=c(1,2))
> boxplot(off$value ~ off$part)
>
> ## when nicely ordre calculation of error is easy
> reference <- rep(c(57.01, 57.06, 57.10, 57.07, 57.12, 57.02, 57.03,
57.08,> 57.01, 57.03), c(9,9,9,9,9, 9,9,9,9,9))
> off$error <- off$value - reference
> boxplot(off$error ~off$part)
Is
off$err<-do.call(c,tapply(off$value,off$part,function(x) x-mean(x)))
what you want to do?
Regards
Petr
>
>
>
>
> This is a constructed example.
>
> How do I find mean for the 10 individual parts and make a off$normalized
> where I substract the mean for the part in the dataset from the
individual> observed values in the data set?
>
> I would also like to find a good method to come from a data.frame/matrix
> where part and reference value and even the avaerage find before where
> listed, the reason is that the order of the data are not always as
> structured:
>
> part reference mean
> 1 57.01 57.0102
> 2 57.06 57.0612
> ....
>
>
> --
> Klaus F. ?stergaard, <farremosen(at)gmail dot com>
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html> and provide commented, minimal, self-contained, reproducible code.