How to use Lines function to draw the error bars?

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
project.org] On Behalf Of liang.che at us.pwc.com
Sent: Tuesday, October 09, 2012 2:50 AM
To: r-help at r-project.org
Subject: Re: [R] How to use Lines function to draw the error bars?

one extra question

How do I export the final image (with plots and lines added) into files
in a batch mode for a bunch of regressions?

thanks


----- Forwarded by Liang Che/US/TLS/PwC on 10/08/2012 08:48 PM -----

From:   Liang Che/US/TLS/PwC
To:     Joseph Clark <joeclark77 at hotmail.com>@INTL
Cc:     r-help at r-project.org
Date:   10/08/2012 08:44 PM
Subject:        RE: [R] How to use Lines function to draw the error
bars?


thank you all for your answers

i have figured out the plot i wanted - it was actually pretty easy

lines(x=20:34,prd[,1],col="red",Ity=1)
lines(x=20:34,prd[,2],lty=2)
lines(x=20:34,prd[,3],lty=2)






From:   Joseph Clark <joeclark77 at hotmail.com>
To:     Liang Che/US/TLS/PwC at Americas-US
Cc:     <r-help at r-project.org>
Date:   10/08/2012 05:30 PM
Subject:        RE: [R] How to use Lines function to draw the error
bars?



I don't have any experience with "predict", but if it returns a data
frame, just capture that data frame and refer to the columns as
vectors.

mydata <- predict(...)
plotCI(  1:nrow(mydata), y=mydata$fit, ui=mydata$upr, li=mydata$lwr )

That seems easier to me.


// joseph w. clark , visiting research associate \\ university of
nebraska at omaha - school of IS&T

To: joeclark77 at hotmail.com
CC: r-help at r-project.org
Subject: RE: [R] How to use Lines function to draw the error bars?
From: liang.che at us.pwc.com
Date: Mon, 8 Oct 2012 17:16:12 -0400

Thanks

since the 'lwr' and 'upr' are produced from the 'predict' function, do
I need to convert the table into a data frame, then define the 'lwr'
and 'upr' as the objects?

upr<-data.frame(prd[,3])
fix(upr)
lwr<-data.frame(prd[,2])
fix(lwr)
y<-data.frame(prd[,1])
fix(y)
plotCI(x=1:15,y=y,uiw=upr,liw=lwr,err=x)


but I got the following error message:

Error in xy.coords(x, y, xlabel, ylabel, log) :
  'x' and 'y' lengths differ
In addition: Warning message:
In if (err == "y") z <- y else z <- x :
  the condition has length > 1 and only the first element will be used




From:        Joseph Clark <joeclark77 at hotmail.com>
To:        Liang Che/US/TLS/PwC at Americas-US
Cc:        <r-help at r-project.org>
Date:        10/08/2012 05:06 PM
Subject:        RE: [R] How to use Lines function to draw the error
bars?



In my example code, 'fit' and 'upr' and 'lwr' are just the names of the
data vectors you gave as an example.  If those names aren't correct,
change them to what you're actually using.  You can also assign your x
values to x.

// joseph w. clark , visiting research associate
\\ university of nebraska at omaha - school of IS&T

To: joeclark77 at hotmail.com
CC: r-help at r-project.org
Subject: RE: [R] How to use Lines function to draw the error bars?
From: liang.che at us.pwc.com
Date: Mon, 8 Oct 2012 17:00:56 -0400

thank you all

plotCI is probably the closest to what I wanted to do -- draw the
Confidence Interval curves around the fitted values (fit)

but I got the following wrong message while I tried plotCI?

Error in plotCI(x = 1:15, y = fit, ui = upr, li = lwr) :
 object 'upr' not found






From:        Joseph Clark <joeclark77 at hotmail.com>
To:        Liang Che/US/TLS/PwC at Americas-US, <r-help at r-project.org>
Date:        10/08/2012 04:03 PM
Subject:        RE: [R] How to use Lines function to draw the error
bars?



I typically use the function "plotCI" from the "plotrix" package for
confidence intervals or error bars.
The code would be:

library(plotrix)
plotCI( x=1:15, y=fit, ui=upr, li=lwr)


// joseph w. clark , visiting research associate
\\ university of nebraska at omaha - school of IS&T

To: r-help at r-project.org
From: liang.che at us.pwc.com
Date: Mon, 8 Oct 2012 15:11:53 -0400
Subject: [R] How to use Lines function to draw the error bars?

fit lwr upr
1 218.4332 90.51019 346.3561
2 218.3906 90.46133 346.3198
3 218.3906 90.46133 346.3198
4 161.3982 44.85702 277.9394
5 192.4450 68.39903 316.4909
6 179.8056 56.49540 303.1158
7 219.5406 91.52707 347.5542
8 162.6761 46.65760 278.6945
9 193.8506 70.59838 317.1029
10 181.3816 58.11305 304.6502
11 221.2871 92.14366 350.4305
12 164.2947 47.91081 280.6785
13 195.3415 72.04109 318.6418
14 182.7447 58.68660 306.8028
15 222.5223 91.86550 353.1791


I have tried

new<-data.frame(newdata$Unemployment)
prd<-predict.lm(fit,newdata,interval=c("confidence"),level=0.95)
lines (new,prd[,3],col="red",lty=2)

but it didn't give me anything.

thanks