This R package based on the paper of Simultaneous Confidence Intervals for Ratios of Means of Log-normal Populations with Zeros by Xu et al. It provides 11 methods for construct simultaneous confidence intervals for ratios of means of LN0. At last, we select 4 excellent methods which based on generalized pivotal quantity with order statistics and two-step MOVER intervals. For the convenience of use, we make a R package called LN0SCIs, and it also has a python version package https://pypi.python.org/pypi/LN0SCIs/0.7.3.
We provaide four main functions in our LN0SCIs packages, GPQW,GPQH,FMW and FMWH, if you want to deep understanding these four methods,you can read our paper: Simultaneous Confidence Intervals for Ratios of Means of Log-normal Populations with Zeros. the code we trust in GitHub. If you want to know how to realize them,you can read the source code.
library(LN0SCIs)
# params setting
p<-c(0.1,0.15,0.1,0.6)
n<-c(30,15,10,50)
mu<-c(1,1.3,2,0)
sigma<-c(1,1,1,2)
C2 <- rbind(c(-1,1,0,0),c(-1,0,1,0),c(-1,0,0,1),c(0,-1,1,0),c(0,-1,0,1),c(0,0,-1,1))
N<-1000
GPQW(n,p,mu,sigma,N,C2 = C2) #base function[1] "====================Method: GPQW===================="
[1] "The Simultaneous Confidence Intervals are: "
[LCL,UCL]
1 [-1.235113,2.848869]
2 [-0.441577,7.030192]
3 [-3.57776,-2.108937]
4 [-1.86122,6.480295]
5 [-5.843864,-1.640372]
6 [-10.008059,-2.477454]
[1] "**********************Time**************************"
Time difference of 53.041 secs
p<-c(0.1,0.15,0.1,0.6)
n<-c(30,15,10,50)
mu<-c(1,1.3,2,0)
sigma<-c(1,1,1,2)
C2 <- rbind(c(-1,1,0,0),c(-1,0,1,0),c(-1,0,0,1),c(0,-1,1,0),c(0,-1,0,1),c(0,0,-1,1))
N<-1000;
GPQH(n,p,mu,sigma,N,C2 = C2)[1] "====================Method: GPQH===================="
[1] "The Simultaneous Confidence Intervals are: "
[LCL,UCL]
1 [-0.996117,1.07758]
2 [-0.36159,2.378184]
3 [-2.87464,1.273622]
4 [-0.302924,2.22018]
5 [-2.954485,1.098756]
6 [-4.086911,0.345388]
[1] "**********************Time**************************"
Time difference of 16.244 secsp <- c(0.1,0.15,0.1,0.6)
n <- c(30,15,10,50)
mu <- c(1,1.3,2,0)
sigma <- c(1,1,1,2)
C2 <- rbind(c(-1,1,0,0),c(-1,0,1,0),c(-1,0,0,1),c(0,-1,1,0),c(0,-1,0,1),c(0,0,-1,1))
N <- 1000
FMW(n,p,mu,sigma,N,C2 = C2)[1] "====================Method: FMW===================="
[1] "The Simultaneous Confidence Intervals are: "
[LCL,UCL]
1 [-1.351672,0.720717]
2 [-1.11909,2.371603]
3 [-1.882617,2.617392]
4 [-0.86023,2.700871]
5 [-1.618718,2.936072]
6 [-2.967078,2.541986]
[1] "**********************Time**************************"
Time difference of 42.577 secsp<-c(0.1,0.15,0.1,0.6)
n<-c(30,15,10,50)
mu<-c(1,1.3,2,0)
sigma<-c(1,1,1,2)
C2 <- rbind(c(-1,1,0,0),c(-1,0,1,0),c(-1,0,0,1),c(0,-1,1,0),c(0,-1,0,1),c(0,0,-1,1))
N<-1000;
FMWH(n,p,mu,sigma,N,C2 = C2)[1] "====================Method: FMWH==================="
[1] "The Simultaneous Confidence Intervals are: "
[LCL,UCL]
1 [-1.334835,1.683172]
2 [-0.806956,2.953145]
3 [-2.683212,1.739062]
4 [-1.130242,2.909716]
5 [-2.979693,1.684558]
6 [-4.145004,1.117301]
[1] "**********************Time**************************"
Time difference of 11.152 secs