> On May 26, 2016, at 7:51 PM, Franco Danilo Roca Landaveri > <[email protected]> wrote: > > Hello, > > I hope you can help me. In class, we were given an Excel worksheet with > specified formulas that take the total score from a survey (or from a > specific section) and convert it to a percentage, according to a table that > assigns scores to a percentile. Since the formulas are too long and > complicated (some have been input by hand) I figured we could fit the data > with a function with some parameters. I plotted the table and sure it > resembled a more-or-less symmetrical quantile function, and I wanted to use R > to find a curve that fitted the data. Here it is: > > percentile <- c(0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.01, 0.05, 0.05, 0.05, > 0.05, 0.05, 0.10, 0.10, 0.15, 0.15, 0.20, 0.25, 0.30, 0.35, 0.40, 0.45, 0.50, > 0.55, 0.60, 0.65, 0.70, 0.75, 0.80, 0.90, 0.95, 0.95, 0.99, 0.99, 0.99, 0.99, > 0.99, 0.99, 0.99, 0.99, 0.99) > > score <- c(10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, > 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, > 44, > 45, 46, 47, 48, 49, 50) >
Use approxfun: myapprox <- approxfun( x=score, y=percentile) > myapprox(27.8) [1] 0.29 > myapprox(27.8) [1] 0.29 > myapprox(50) [1] 0.99 > myapprox(51) [1] NA You can change how values outside the max and min limits are handles with furhter arguments to approxfun. — David. > I looked up the quantreg package, but I didn't know how to use it properly > since I don't have the raw data, only the percentiles, and also because what > I'm trying to get is the percentile based on the score, not the other way > around. Then I tried to fit it using a sigmoid curve (similar to a cdf), > using the following code: > > require(drc) > model1 <- drm(percentile ~ score, fct = LL.4()) > summary(model1) > > But I found another problem, the fitted curve apparently went over 1 on the y > axis, something not possible for a cdf. I'd like to know what would be the > most appropiate way to do this, and also if it is possible to fit a quantile > function with this data, apart from the cdf. > > Thank you very much for your help > > ______________________________________________ > [email protected] mailing list -- To UNSUBSCRIBE and more, see > 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. ______________________________________________ [email protected] mailing list -- To UNSUBSCRIBE and more, see 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.
