Hello,
I have a question regarding fitting a polynomial to a data set, then
constructing a polynom from the coefficients so that I can integrate it. I
first use lm to fit the polynomial setting the coefficients to raw=TRUE -
this appears to work fine. I plot the model and it is a true
representation of the data.
I then take the coefficients vector and construct a polynom from the
library(PolynomF) package. When I plot this the generated line is not
representative of the data. The polynomial intercept is correct, however
the gradient is well below the actual dataset, and does not follow the
same contour.
A code snippet and output are provided below. Any help would be much
appreciated.
Thanks,
Blaise.
lddata.mdl <- lm(lddata[,2]~poly(lddata[,1],9,raw=TRUE))
lddata.ply <- polynom(lddata.mdl$coefficients);
lddata.int <- integral(lddata.ply);
plot(lddata[,1],lddata[,2],pch=".")
lines(predict(lddata.ply,timeseq),col=4,lwd=2,lty=1)
lines(predict(lddata.mdl,newdata=list(timeseq=x)),col=4,lwd=2,lty=1)
The coefficients from lddata.mdl are:
(Intercept) poly(lddata[, 1], 9, raw = TRUE)1
-4.714865e+05 1.054330e+06
poly(lddata[, 1], 9, raw = TRUE)2 poly(lddata[, 1], 9, raw = TRUE)3
-1.855263e+05 2.397431e+04
poly(lddata[, 1], 9, raw = TRUE)4 poly(lddata[, 1], 9, raw = TRUE)5
-1.692492e+03 6.492181e+01
poly(lddata[, 1], 9, raw = TRUE)6 poly(lddata[, 1], 9, raw = TRUE)7
-1.335839e+00 1.393105e-02
poly(lddata[, 1], 9, raw = TRUE)8 poly(lddata[, 1], 9, raw = TRUE)9
-7.284687e-05 3.200856e-07
The output from the lddata.ply is:
-471486.5 + 1054330*x - 185526.3*x^2 + 23974.31*x^3 - 1692.492*x^4 +
64.92181*x^5 - 1.335839*x^6 + 0.01393105*x^7 - 7.284687e-05*x^8 +
3.200856e-07*x^9
---
This e-mail may contain confidential and/or privileged i...{{dropped:8}}
______________________________________________
R-help@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.
