"R_frb" not available for .C() for package "FRB"

[guhjy]

a <- lmrob(LNOx ~ LNOxEm + sqrtWS, data=NOxEmissions)
set.seed(123)
tmp <- frb(lmrob.object=a, nboot=1000, return.coef=FALSE)
Error in .C("R_frb", as.double(xx), as.double(yy), as.double(w), as.integer(n), :
"R_frb" not available for .C() for package "FRB"

[msalibian]

@guhjy I can't reproduce the error. I just re-installed FRB and ran the above without problems (output below). Did "install_github" build the package properly when your last ran it?

library(devtools)
install_github("msalibian/FRB")
library(FRB)
library(robustbase)
data(NOxEmissions)
a <- lmrob(LNOx ~ LNOxEm + sqrtWS, data=NOxEmissions)
set.seed(123)
( tmp <- frb(lmrob.object=a, nboot=1000, return.coef=FALSE) )
              [,1]          [,2]          [,3]
[1,]  0.0029529150 -3.911259e-04  6.996540e-05
[2,] -0.0003911259  5.827418e-05 -4.159314e-05
[3,]  0.0000699654 -4.159314e-05  1.786090e-04

[guhjy]

@msalibian: The problem has been solved. Thank you! However, this installation erases the CRAN FRB package ( https://cran.r-project.org/web/packages/FRB/index.html), which has the same name as your package. 1. What to do if two packages have the same name? 2. What are the differences between your package and the CRAN "FRB" package? 3. How to compare the coefficients between 2 lmrob models? e.g.

((a <- lmrob(mpg ~ disp + hp + wt, data=mtcars)))

Call: lmrob(formula = mpg ~ disp + hp + wt, data = mtcars) \--> method = "MM" Coefficients: (Intercept) disp hp wt 36.6321312 0.0009983 -0.0303098 -3.9208073

((b <- lmrob(mpg ~ disp + hp + qsec, data=mtcars)))

Call: lmrob(formula = mpg ~ disp + hp + qsec, data = mtcars) \--> method = "MM" Coefficients: (Intercept) disp hp qsec 38.02182 -0.02632 -0.03292 -0.41532

((tmp1 <- frb(lmrob.object=a, nboot=1000, return.coef=FALSE)))

[,1] [,2] [,3] [,4] [1,] 9.96226291 0.0162143443 -0.0203900060 -3.061000146 [2,] 0.01621434 0.0000776010 -0.0000734849 -0.006770139 [3,] -0.02039001 -0.0000734849 0.0001533673 0.003630823 [4,] -3.06100015 -0.0067701395 0.0036308232 1.224974120

((tmp2 <- frb(lmrob.object=b, nboot=1000, return.coef=FALSE)))

[,1] [,2] [,3] [,4] [1,] 80.06564391 0.0260099385 -0.1202609246 -3.846005301 [2,] 0.02600994 0.0001865492 -0.0001639225 -0.003102574 [3,] -0.12026092 -0.0001639225 0.0003585570 0.006195492 [4,] -3.84600530 -0.0031025742 0.0061954923 0.213361260

((tmp3 <- frb(lmrob.object=coef(a)-coef(b), nboot=1000, return.coef=FALSE)))

Error: $ operator is invalid for atomic vectors Thank you. Sincerely Yours, Jinn-Yuh Guh, M.D. Division of Nephrology Department of Internal Medicine Kaohsiung Medical University 100 Zihyou 1st Rd. Kaohsiung, Taiwan 80756 E-mail:[email protected] TEL: 886-7-3121101 EXT.7353~12 FAX: 886-7-3228721 2017-10-22 10:16 GMT+08:00 Matias Salibian Barrera <[email protected]

…

: @guhjy <https://github.com/guhjy> I can't reproduce the error. I just re-installed FRB and ran the above without problems (output below). Did "install_github" build the package properly when your last ran it? library(devtools) install_github("msalibian/FRB") library(FRB) library(robustbase) data(NOxEmissions) a <- lmrob(LNOx ~ LNOxEm + sqrtWS, data=NOxEmissions) set.seed(123) ( tmp <- frb(lmrob.object=a, nboot=1000, return.coef=FALSE) ) [,1] [,2] [,3] [1,] 0.0029529150 -3.911259e-04 6.996540e-05 [2,] -0.0003911259 5.827418e-05 -4.159314e-05 [3,] 0.0000699654 -4.159314e-05 1.786090e-04 — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#1 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AA6M52Y95xow4oGYvHuAiWyPkCcIrR8_ks5suqVvgaJpZM4QBm24> .

[msalibian]

1. What to do if two packages have the same name?
   Thanks for pointing this out. I think I will need to rename my package. Maybe to FRB.reg, or something that reflects the fact that my package only deals with linear regression models.

2. What are the differences between your package and the CRAN "FRB" package?
   The methodology is the same. My package deals with (scalar-response) linear regression models, the other one deals with multivariate models.