Lorenzo Isella
2016-Mar-19 20:28 UTC
[R] Orthogonal Nonlinear Least-Squares Regression in R
Dear All,
I am trying my hands at orthogonal least square regression.
Have a look for instance at
http://bit.ly/1pB2aHX
https://cran.r-project.org/web/packages/onls/index.html
http://bit.ly/1XDkkTL
docs.scipy.org/doc/external/odrpack_guide.pdf
However, I am experiencing some problems with a simple example, just
to compare the results to a simple linear fit.
Have a look at the following snippet
################################################
library(onls)
dd <- structure(list(de_emp_mn_agg = c(9226, 9404, 9604, 10183, 10788,
11352, 11984, 12921, 14057, 15235, 15560, 15738, 16039, 16729,
17332, 18398, 19458, 20001, 19861, 20690, 21495, 21869, 22145,
22521), de_emp_ind_agg = c(43862, 40621, 37884, 36039, 35228,
34336, 33684, 33816, 33593, 33861, 33817, 33139, 32250, 31796,
31276, 30934, 31340, 32078, 31366, 30800, 31410, 31975, 32120,
32254), de_gov_exp_agg = c(1183695, 1281056, 1334560, 1390475,
1439175, 1472564, 1495661, 1520460, 1565451, 1604454, 1654996,
1672559, 1701664, 1722003, 1751547, 1793235, 1824640, 1874300,
1894248, 1939610, 2001224, 2056539, 2104642, 2156210), berd c(26245.5,
26579, 25933, 25910, 26816.6, 27211, 28909.8, 30334.44, 33622.55,
35600, 36331.9, 36950, 38029, 38363, 38651.038, 41148, 43034,
46073, 45275, 46929, 51077.2, 53790.1, 53566.2, 56226)), row.names c(NA,
24L), .Names = c("de_emp_mn_agg", "de_emp_ind_agg",
"de_gov_exp_agg",
"berd"), class = "data.frame")
mm1 <- lm(berd~ de_emp_mn_agg+
de_emp_ind_agg+
de_gov_exp_agg ,
data=dd)
#and now this fails
mm3 <- onls(berd~ K+ A*de_emp_mn_agg+
B*de_emp_ind_agg+
D*de_gov_exp_agg,
data=dd,
start = list(K=-5e4, A= 1,B =1, D=0.1 ))
################################################
I get this error message
Error in mf[[varNamesRHS]] : recursive indexing failed at level 2
What am I doing wrong?
Here is my sessionInfo()
> sessionInfo()
R version 3.2.4 Revised (2016-03-16 r70336)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Debian GNU/Linux 8 (jessie)
locale:
[1] LC_CTYPE=en_GB.utf8 LC_NUMERIC=C
[3] LC_TIME=en_GB.utf8 LC_COLLATE=en_GB.utf8
[5] LC_MONETARY=en_GB.utf8 LC_MESSAGES=en_GB.utf8
[7] LC_PAPER=en_GB.utf8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_GB.utf8 LC_IDENTIFICATION=C
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] onls_0.1-1 minpack.lm_1.2-0
Any suggestion is appreciated.
Many thanks
Lorenzo