

Hi R User,
I was trying to calculate ratios with confidence interval using Monte Carlo
simulation but I could not figure it out.
Here is the example of my data (see below), I want to calculate ratios
(dat$v1/dat$v3 & dat$v2/dat$v3) and its confidence intervals using a 100
randomly selected data sets.
Could you please give me your suggestions how I can estimate ratios with
CI?
I will be very grateful to you.
Sincerely,
MW

dat<structure(list(v1 = c(NA, TRUE, TRUE, TRUE, TRUE, TRUE, NA, TRUE,
NA, NA, TRUE, TRUE, TRUE, TRUE, NA, NA, TRUE, TRUE), v2 = c(TRUE,
NA, NA, NA, NA, TRUE, NA, NA, TRUE, TRUE, NA, TRUE, TRUE, NA,
NA, TRUE, TRUE, NA), v3 = c(TRUE, TRUE, NA, TRUE, TRUE, NA, NA,
TRUE, TRUE, NA, NA, TRUE, TRUE, TRUE, NA, NA, TRUE, NA)), .Names = c("v1",
"v2", "v3"), class = "data.frame", row.names = c(NA, 18L))
ratio1<length(which(dat$v1 == "TRUE"))/length(which(dat$v3 == "TRUE"))
ratio2<length(which(dat$v2 == "TRUE"))/length(which(dat$v3 == "TRUE"))
Thanks
[[alternative HTML version deleted]]
______________________________________________
[hidden email] mailing list  To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/rhelpPLEASE do read the posting guide http://www.Rproject.org/postingguide.htmland provide commented, minimal, selfcontained, reproducible code.


> ratio1 < with(dat, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
> ratio1
[1] 1.2
It looks like you should spend some more time with an R tutorial or two.
This is basic stuff (if I understand what you wanted correctly).
Also, this is not how a "confidence interval" should be calculated, but
that is another off topic discussion for which stats.stackexchange.com is a
more appropriate venue.
Cheers,
Bert
Bert Gunter
"The trouble with having an open mind is that people keep coming along and
sticking things into it."
 Opus (aka Berkeley Breathed in his "Bloom County" comic strip )
On Mon, Mar 25, 2019 at 4:31 PM Marna Wagley < [hidden email]> wrote:
> Hi R User,
> I was trying to calculate ratios with confidence interval using Monte Carlo
> simulation but I could not figure it out.
> Here is the example of my data (see below), I want to calculate ratios
> (dat$v1/dat$v3 & dat$v2/dat$v3) and its confidence intervals using a 100
> randomly selected data sets.
> Could you please give me your suggestions how I can estimate ratios with
> CI?
> I will be very grateful to you.
> Sincerely,
>
> MW
> 
> dat<structure(list(v1 = c(NA, TRUE, TRUE, TRUE, TRUE, TRUE, NA, TRUE,
>
> NA, NA, TRUE, TRUE, TRUE, TRUE, NA, NA, TRUE, TRUE), v2 = c(TRUE,
>
> NA, NA, NA, NA, TRUE, NA, NA, TRUE, TRUE, NA, TRUE, TRUE, NA,
>
> NA, TRUE, TRUE, NA), v3 = c(TRUE, TRUE, NA, TRUE, TRUE, NA, NA,
>
> TRUE, TRUE, NA, NA, TRUE, TRUE, TRUE, NA, NA, TRUE, NA)), .Names = c("v1",
>
> "v2", "v3"), class = "data.frame", row.names = c(NA, 18L))
>
>
> ratio1<length(which(dat$v1 == "TRUE"))/length(which(dat$v3 == "TRUE"))
>
> ratio2<length(which(dat$v2 == "TRUE"))/length(which(dat$v3 == "TRUE"))
>
>
> Thanks
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> [hidden email] mailing list  To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/rhelp> PLEASE do read the posting guide
> http://www.Rproject.org/postingguide.html> and provide commented, minimal, selfcontained, reproducible code.
>
[[alternative HTML version deleted]]
______________________________________________
[hidden email] mailing list  To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/rhelpPLEASE do read the posting guide http://www.Rproject.org/postingguide.htmland provide commented, minimal, selfcontained, reproducible code.


Dear Bert,
Thank you very much for the response.
I did it manually but I could not put them in a loop so that I created the
table manually with selecting the rows randomly several times. Here what I
have done so far, please find it. I want to create the table 100 times and
calculate its mean and CI from those 100 values. If anyone can give me some
hint to make a loop, that would be great. I am very grateful with your help.
Thanks,
library(dplyr)
library(plyr)
dat<structure(list(v1 = c(NA, TRUE, TRUE, TRUE, TRUE, TRUE, NA, TRUE,
NA, NA, TRUE, TRUE, TRUE, TRUE, NA, NA, TRUE, TRUE), v2 = c(TRUE,
NA, NA, NA, NA, TRUE, NA, NA, TRUE, TRUE, NA, TRUE, TRUE, NA,
NA, TRUE, TRUE, NA), v3 = c(TRUE, TRUE, NA, TRUE, TRUE, NA, NA,
TRUE, TRUE, NA, NA, TRUE, TRUE, TRUE, NA, NA, TRUE, NA)), .Names = c("v1",
"v2", "v3"), class = "data.frame", row.names = c(NA, 18L))
ratio1 < with(dat, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
ratio2 < with(dat, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
#
A1<sample_n(dat1, 16)# created a table with selecting a 16 sample size
(rows)
A1.ratio1<with(A1, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
A1.ratio2 < with(A1, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
A1.Table<data.frame(Ratio1=A1.ratio1, Ratio2=A1.ratio2)
#
A2<sample_n(dat1, 16)
A2.ratio1<with(A2, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
A2.ratio2 < with(A2, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
A2.Table<data.frame(Ratio1=A2.ratio1, Ratio2=A2.ratio2)
#
A3<sample_n(dat1, 16)
A3.ratio1<with(A3, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
A3.ratio2 < with(A3, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
A3.Table<data.frame(Ratio1=A3.ratio1, Ratio2=A3.ratio2)
#
##..............
# I was thinking to repeat this procedure 100 times and calculate the ratio
A100<sample_n(dat1, 16)
A100.ratio1<with(A100, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
A100.ratio2 < with(A100, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
A100.Table<data.frame(Ratio1=A100.ratio1, Ratio2=A100.ratio2)
#
Tab<rbind(A1.Table, A2.Table, A3.Table, A100.Table)
#Compute the mean for each ratio
Ratio1<mean(Table1[,1])
Ratio2<mean(Table1[,2])
summary < ddply(subset(Tab), c(""),summarise,
N = length(Tab),
mean.R1 = mean(Ratio1, na.rm=T),
median.R1=median(Ratio1, na.rm=T),
sd.R1 = sd(Ratio1, na.rm=T),
se.R1 = sd / sqrt(N),
LCI.95.R1=mean.R11.95*se.R1,
UCI.95.R1=mean.R1+1.95*se.R1,
mean.R2 = mean(Ratio2, na.rm=T),
median.R2=median(Ratio2, na.rm=T),
sd.R2 = sd(Ratio2, na.rm=T),
se.R2 = sd / sqrt(N),
LCI.95.R2=mean.R21.95*se.R2,
UCI.95.R2=mean.R2+1.95*se.R2
)
summary
On Mon, Mar 25, 2019 at 4:50 PM Bert Gunter < [hidden email]> wrote:
>
> > ratio1 < with(dat, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
> > ratio1
> [1] 1.2
>
> It looks like you should spend some more time with an R tutorial or two.
> This is basic stuff (if I understand what you wanted correctly).
>
> Also, this is not how a "confidence interval" should be calculated, but
> that is another off topic discussion for which stats.stackexchange.com is
> a more appropriate venue.
>
> Cheers,
> Bert
>
> Bert Gunter
>
> "The trouble with having an open mind is that people keep coming along and
> sticking things into it."
>  Opus (aka Berkeley Breathed in his "Bloom County" comic strip )
>
>
> On Mon, Mar 25, 2019 at 4:31 PM Marna Wagley < [hidden email]>
> wrote:
>
>> Hi R User,
>> I was trying to calculate ratios with confidence interval using Monte
>> Carlo
>> simulation but I could not figure it out.
>> Here is the example of my data (see below), I want to calculate ratios
>> (dat$v1/dat$v3 & dat$v2/dat$v3) and its confidence intervals using a 100
>> randomly selected data sets.
>> Could you please give me your suggestions how I can estimate ratios with
>> CI?
>> I will be very grateful to you.
>> Sincerely,
>>
>> MW
>> 
>> dat<structure(list(v1 = c(NA, TRUE, TRUE, TRUE, TRUE, TRUE, NA, TRUE,
>>
>> NA, NA, TRUE, TRUE, TRUE, TRUE, NA, NA, TRUE, TRUE), v2 = c(TRUE,
>>
>> NA, NA, NA, NA, TRUE, NA, NA, TRUE, TRUE, NA, TRUE, TRUE, NA,
>>
>> NA, TRUE, TRUE, NA), v3 = c(TRUE, TRUE, NA, TRUE, TRUE, NA, NA,
>>
>> TRUE, TRUE, NA, NA, TRUE, TRUE, TRUE, NA, NA, TRUE, NA)), .Names = c("v1",
>>
>> "v2", "v3"), class = "data.frame", row.names = c(NA, 18L))
>>
>>
>> ratio1<length(which(dat$v1 == "TRUE"))/length(which(dat$v3 == "TRUE"))
>>
>> ratio2<length(which(dat$v2 == "TRUE"))/length(which(dat$v3 == "TRUE"))
>>
>>
>> Thanks
>>
>> [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> [hidden email] mailing list  To UNSUBSCRIBE and more, see
>> https://stat.ethz.ch/mailman/listinfo/rhelp>> PLEASE do read the posting guide
>> http://www.Rproject.org/postingguide.html>> and provide commented, minimal, selfcontained, reproducible code.
>>
>
[[alternative HTML version deleted]]
______________________________________________
[hidden email] mailing list  To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/rhelpPLEASE do read the posting guide http://www.Rproject.org/postingguide.htmland provide commented, minimal, selfcontained, reproducible code.


Do you really not know how to use a for loop? The tutorial recommendation seems apropos...
On March 26, 2019 5:57:17 AM PDT, Marna Wagley < [hidden email]> wrote:
>Dear Bert,
>Thank you very much for the response.
>I did it manually but I could not put them in a loop so that I created
>the
>table manually with selecting the rows randomly several times. Here
>what I
>have done so far, please find it. I want to create the table 100 times
>and
>calculate its mean and CI from those 100 values. If anyone can give me
>some
>hint to make a loop, that would be great. I am very grateful with your
>help.
>Thanks,
>
>library(dplyr)
>
>library(plyr)
>
>dat<structure(list(v1 = c(NA, TRUE, TRUE, TRUE, TRUE, TRUE, NA, TRUE,
>
>NA, NA, TRUE, TRUE, TRUE, TRUE, NA, NA, TRUE, TRUE), v2 = c(TRUE,
>
>NA, NA, NA, NA, TRUE, NA, NA, TRUE, TRUE, NA, TRUE, TRUE, NA,
>
>NA, TRUE, TRUE, NA), v3 = c(TRUE, TRUE, NA, TRUE, TRUE, NA, NA,
>
>TRUE, TRUE, NA, NA, TRUE, TRUE, TRUE, NA, NA, TRUE, NA)), .Names =
>c("v1",
>
>"v2", "v3"), class = "data.frame", row.names = c(NA, 18L))
>
>
>ratio1 < with(dat, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>ratio2 < with(dat, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>#
>
>A1<sample_n(dat1, 16)# created a table with selecting a 16 sample size
>(rows)
>
>A1.ratio1<with(A1, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A1.ratio2 < with(A1, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A1.Table<data.frame(Ratio1=A1.ratio1, Ratio2=A1.ratio2)
>
>#
>
>A2<sample_n(dat1, 16)
>
>A2.ratio1<with(A2, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A2.ratio2 < with(A2, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A2.Table<data.frame(Ratio1=A2.ratio1, Ratio2=A2.ratio2)
>
>#
>
>A3<sample_n(dat1, 16)
>
>A3.ratio1<with(A3, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A3.ratio2 < with(A3, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A3.Table<data.frame(Ratio1=A3.ratio1, Ratio2=A3.ratio2)
>
>#
>
>##..............
>
># I was thinking to repeat this procedure 100 times and calculate the
>ratio
>
>A100<sample_n(dat1, 16)
>
>A100.ratio1<with(A100, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A100.ratio2 < with(A100, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A100.Table<data.frame(Ratio1=A100.ratio1, Ratio2=A100.ratio2)
>
>#
>
>Tab<rbind(A1.Table, A2.Table, A3.Table, A100.Table)
>
>
>#Compute the mean for each ratio
>
>Ratio1<mean(Table1[,1])
>
>Ratio2<mean(Table1[,2])
>
>
>summary < ddply(subset(Tab), c(""),summarise,
>
> N = length(Tab),
>
> mean.R1 = mean(Ratio1, na.rm=T),
>
> median.R1=median(Ratio1, na.rm=T),
>
> sd.R1 = sd(Ratio1, na.rm=T),
>
> se.R1 = sd / sqrt(N),
>
> LCI.95.R1=mean.R11.95*se.R1,
>
> UCI.95.R1=mean.R1+1.95*se.R1,
>
>
>
> mean.R2 = mean(Ratio2, na.rm=T),
>
> median.R2=median(Ratio2, na.rm=T),
>
> sd.R2 = sd(Ratio2, na.rm=T),
>
> se.R2 = sd / sqrt(N),
>
> LCI.95.R2=mean.R21.95*se.R2,
>
> UCI.95.R2=mean.R2+1.95*se.R2
>
> )
>
> summary
>
>
>
>On Mon, Mar 25, 2019 at 4:50 PM Bert Gunter < [hidden email]>
>wrote:
>
>>
>> > ratio1 < with(dat, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>> > ratio1
>> [1] 1.2
>>
>> It looks like you should spend some more time with an R tutorial or
>two.
>> This is basic stuff (if I understand what you wanted correctly).
>>
>> Also, this is not how a "confidence interval" should be calculated,
>but
>> that is another off topic discussion for which
>stats.stackexchange.com is
>> a more appropriate venue.
>>
>> Cheers,
>> Bert
>>
>> Bert Gunter
>>
>> "The trouble with having an open mind is that people keep coming
>along and
>> sticking things into it."
>>  Opus (aka Berkeley Breathed in his "Bloom County" comic strip )
>>
>>
>> On Mon, Mar 25, 2019 at 4:31 PM Marna Wagley < [hidden email]>
>> wrote:
>>
>>> Hi R User,
>>> I was trying to calculate ratios with confidence interval using
>Monte
>>> Carlo
>>> simulation but I could not figure it out.
>>> Here is the example of my data (see below), I want to calculate
>ratios
>>> (dat$v1/dat$v3 & dat$v2/dat$v3) and its confidence intervals using a
>100
>>> randomly selected data sets.
>>> Could you please give me your suggestions how I can estimate ratios
>with
>>> CI?
>>> I will be very grateful to you.
>>> Sincerely,
>>>
>>> MW
>>> 
>>> dat<structure(list(v1 = c(NA, TRUE, TRUE, TRUE, TRUE, TRUE, NA,
>TRUE,
>>>
>>> NA, NA, TRUE, TRUE, TRUE, TRUE, NA, NA, TRUE, TRUE), v2 = c(TRUE,
>>>
>>> NA, NA, NA, NA, TRUE, NA, NA, TRUE, TRUE, NA, TRUE, TRUE, NA,
>>>
>>> NA, TRUE, TRUE, NA), v3 = c(TRUE, TRUE, NA, TRUE, TRUE, NA, NA,
>>>
>>> TRUE, TRUE, NA, NA, TRUE, TRUE, TRUE, NA, NA, TRUE, NA)), .Names =
>c("v1",
>>>
>>> "v2", "v3"), class = "data.frame", row.names = c(NA, 18L))
>>>
>>>
>>> ratio1<length(which(dat$v1 == "TRUE"))/length(which(dat$v3 ==
>"TRUE"))
>>>
>>> ratio2<length(which(dat$v2 == "TRUE"))/length(which(dat$v3 ==
>"TRUE"))
>>>
>>>
>>> Thanks
>>>
>>> [[alternative HTML version deleted]]
>>>
>>> ______________________________________________
>>> [hidden email] mailing list  To UNSUBSCRIBE and more, see
>>> https://stat.ethz.ch/mailman/listinfo/rhelp>>> PLEASE do read the posting guide
>>> http://www.Rproject.org/postingguide.html>>> and provide commented, minimal, selfcontained, reproducible code.
>>>
>>
>
> [[alternative HTML version deleted]]
>
>______________________________________________
> [hidden email] mailing list  To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/rhelp>PLEASE do read the posting guide
> http://www.Rproject.org/postingguide.html>and provide commented, minimal, selfcontained, reproducible code.

Sent from my phone. Please excuse my brevity.
______________________________________________
[hidden email] mailing list  To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/rhelpPLEASE do read the posting guide http://www.Rproject.org/postingguide.htmland provide commented, minimal, selfcontained, reproducible code.


I second the vote on needing a tutorial. You need to learn about how R does things and get familiar with vectorization and the apply() family of functions. You defined dat but not dat1 in your code so I'll just use dat. First, to get the ratios:
(ratios < colSums(dat[3], na.rm=TRUE)/colSums(dat[3], na.rm=TRUE))
# v1 v2
# 1.2 0.8
Then create a function for the Monte Carlo simulation that generates a sample and computes the ratios. Finally, use the function with replicate() to generate the 100 samples:
nratios < function(x) {
sdat < x[sample.int(18,16), ]
colSums(sdat[3], na.rm=TRUE)/colSums(sdat[3], na.rm=TRUE)
}
mcrat < replicate(100, nratios(dat))
str(mcrat)
# num [1:2, 1:100] 1 0.8 1.222 0.778 1.111 ...
#  attr(*, "dimnames")=List of 2
# ..$ : chr [1:2] "v1" "v2"
# ..$ : NULL
100 values of ratio1 are stored as mcrat["v1", ] and 100 values of ratio2 are stored as mcrat["v2", ].
Now you can generate your summary statistics.

David L Carlson
Department of Anthropology
Texas A&M University
College Station, TX 778434352
Original Message
From: Rhelp < [hidden email]> On Behalf Of Jeff Newmiller
Sent: Tuesday, March 26, 2019 9:27 AM
To: [hidden email]; Marna Wagley < [hidden email]>; Bert Gunter < [hidden email]>
Cc: rhelp mailing list < [hidden email]>
Subject: Re: [R] Monte Carlo simulation for ratio and its CI
Do you really not know how to use a for loop? The tutorial recommendation seems apropos...
On March 26, 2019 5:57:17 AM PDT, Marna Wagley < [hidden email]> wrote:
>Dear Bert,
>Thank you very much for the response.
>I did it manually but I could not put them in a loop so that I created
>the
>table manually with selecting the rows randomly several times. Here
>what I
>have done so far, please find it. I want to create the table 100 times
>and
>calculate its mean and CI from those 100 values. If anyone can give me
>some
>hint to make a loop, that would be great. I am very grateful with your
>help.
>Thanks,
>
>library(dplyr)
>
>library(plyr)
>
>dat<structure(list(v1 = c(NA, TRUE, TRUE, TRUE, TRUE, TRUE, NA, TRUE,
>
>NA, NA, TRUE, TRUE, TRUE, TRUE, NA, NA, TRUE, TRUE), v2 = c(TRUE,
>
>NA, NA, NA, NA, TRUE, NA, NA, TRUE, TRUE, NA, TRUE, TRUE, NA,
>
>NA, TRUE, TRUE, NA), v3 = c(TRUE, TRUE, NA, TRUE, TRUE, NA, NA,
>
>TRUE, TRUE, NA, NA, TRUE, TRUE, TRUE, NA, NA, TRUE, NA)), .Names =
>c("v1",
>
>"v2", "v3"), class = "data.frame", row.names = c(NA, 18L))
>
>
>ratio1 < with(dat, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>ratio2 < with(dat, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>#
>
>A1<sample_n(dat1, 16)# created a table with selecting a 16 sample size
>(rows)
>
>A1.ratio1<with(A1, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A1.ratio2 < with(A1, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A1.Table<data.frame(Ratio1=A1.ratio1, Ratio2=A1.ratio2)
>
>#
>
>A2<sample_n(dat1, 16)
>
>A2.ratio1<with(A2, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A2.ratio2 < with(A2, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A2.Table<data.frame(Ratio1=A2.ratio1, Ratio2=A2.ratio2)
>
>#
>
>A3<sample_n(dat1, 16)
>
>A3.ratio1<with(A3, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A3.ratio2 < with(A3, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A3.Table<data.frame(Ratio1=A3.ratio1, Ratio2=A3.ratio2)
>
>#
>
>##..............
>
># I was thinking to repeat this procedure 100 times and calculate the
>ratio
>
>A100<sample_n(dat1, 16)
>
>A100.ratio1<with(A100, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A100.ratio2 < with(A100, sum(v2,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>
>A100.Table<data.frame(Ratio1=A100.ratio1, Ratio2=A100.ratio2)
>
>#
>
>Tab<rbind(A1.Table, A2.Table, A3.Table, A100.Table)
>
>
>#Compute the mean for each ratio
>
>Ratio1<mean(Table1[,1])
>
>Ratio2<mean(Table1[,2])
>
>
>summary < ddply(subset(Tab), c(""),summarise,
>
> N = length(Tab),
>
> mean.R1 = mean(Ratio1, na.rm=T),
>
> median.R1=median(Ratio1, na.rm=T),
>
> sd.R1 = sd(Ratio1, na.rm=T),
>
> se.R1 = sd / sqrt(N),
>
> LCI.95.R1=mean.R11.95*se.R1,
>
> UCI.95.R1=mean.R1+1.95*se.R1,
>
>
>
> mean.R2 = mean(Ratio2, na.rm=T),
>
> median.R2=median(Ratio2, na.rm=T),
>
> sd.R2 = sd(Ratio2, na.rm=T),
>
> se.R2 = sd / sqrt(N),
>
> LCI.95.R2=mean.R21.95*se.R2,
>
> UCI.95.R2=mean.R2+1.95*se.R2
>
> )
>
> summary
>
>
>
>On Mon, Mar 25, 2019 at 4:50 PM Bert Gunter < [hidden email]>
>wrote:
>
>>
>> > ratio1 < with(dat, sum(v1,na.rm = TRUE)/sum(v3,na.rm=TRUE))
>> > ratio1
>> [1] 1.2
>>
>> It looks like you should spend some more time with an R tutorial or
>two.
>> This is basic stuff (if I understand what you wanted correctly).
>>
>> Also, this is not how a "confidence interval" should be calculated,
>but
>> that is another off topic discussion for which
>stats.stackexchange.com is
>> a more appropriate venue.
>>
>> Cheers,
>> Bert
>>
>> Bert Gunter
>>
>> "The trouble with having an open mind is that people keep coming
>along and
>> sticking things into it."
>>  Opus (aka Berkeley Breathed in his "Bloom County" comic strip )
>>
>>
>> On Mon, Mar 25, 2019 at 4:31 PM Marna Wagley < [hidden email]>
>> wrote:
>>
>>> Hi R User,
>>> I was trying to calculate ratios with confidence interval using
>Monte
>>> Carlo
>>> simulation but I could not figure it out.
>>> Here is the example of my data (see below), I want to calculate
>ratios
>>> (dat$v1/dat$v3 & dat$v2/dat$v3) and its confidence intervals using a
>100
>>> randomly selected data sets.
>>> Could you please give me your suggestions how I can estimate ratios
>with
>>> CI?
>>> I will be very grateful to you.
>>> Sincerely,
>>>
>>> MW
>>> 
>>> dat<structure(list(v1 = c(NA, TRUE, TRUE, TRUE, TRUE, TRUE, NA,
>TRUE,
>>>
>>> NA, NA, TRUE, TRUE, TRUE, TRUE, NA, NA, TRUE, TRUE), v2 = c(TRUE,
>>>
>>> NA, NA, NA, NA, TRUE, NA, NA, TRUE, TRUE, NA, TRUE, TRUE, NA,
>>>
>>> NA, TRUE, TRUE, NA), v3 = c(TRUE, TRUE, NA, TRUE, TRUE, NA, NA,
>>>
>>> TRUE, TRUE, NA, NA, TRUE, TRUE, TRUE, NA, NA, TRUE, NA)), .Names =
>c("v1",
>>>
>>> "v2", "v3"), class = "data.frame", row.names = c(NA, 18L))
>>>
>>>
>>> ratio1<length(which(dat$v1 == "TRUE"))/length(which(dat$v3 ==
>"TRUE"))
>>>
>>> ratio2<length(which(dat$v2 == "TRUE"))/length(which(dat$v3 ==
>"TRUE"))
>>>
>>>
>>> Thanks
>>>
>>> [[alternative HTML version deleted]]
>>>
>>> ______________________________________________
>>> [hidden email] mailing list  To UNSUBSCRIBE and more, see
>>> https://stat.ethz.ch/mailman/listinfo/rhelp>>> PLEASE do read the posting guide
>>> http://www.Rproject.org/postingguide.html>>> and provide commented, minimal, selfcontained, reproducible code.
>>>
>>
>
> [[alternative HTML version deleted]]
>
>______________________________________________
> [hidden email] mailing list  To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/rhelp>PLEASE do read the posting guide
> http://www.Rproject.org/postingguide.html>and provide commented, minimal, selfcontained, reproducible code.

Sent from my phone. Please excuse my brevity.
______________________________________________
[hidden email] mailing list  To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/rhelpPLEASE do read the posting guide http://www.Rproject.org/postingguide.htmland provide commented, minimal, selfcontained, reproducible code.
______________________________________________
[hidden email] mailing list  To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/rhelpPLEASE do read the posting guide http://www.Rproject.org/postingguide.htmland provide commented, minimal, selfcontained, reproducible code.

