Hi:
Your confidence intervals are so short that the size of the point in the
graphics region covers the endpoints! You also have a wide range of
simulated means (0 - 52) and actual values (0 - 54). Here are some measures
of your CIs:
> with(simvsact, max(simCI.upper - simCI.lower)) # maximum CI width
[1] 0.02346451
> with(simvsact, min(simCI.upper - simCI.lower)) # minimum
[1] 0.001726888
> with(simvsact, mean(simCI.upper - simCI.lower)) # average
[1] 0.00571329
On a y-scale of 0 - 55, say, can you see that the CI widths are a problem?
You may as well just plot the means and actual values, which is easy enough
to do in any of the graphics engines.
HTH,
Dennis
On Wed, Oct 20, 2010 at 2:41 AM, Peter Francis <peterfran...@me.com> wrote:
> Dear List,
>
> I am relatively new to R and am trying to create more attractive plots than
> excel can manage!
>
> I have looked through the various programmes ggplot, lattice, hmisc etc but
> my case seems to be not metnioned, maybe it is but i have not noticed - if
> this is the case i apologise.
>
>
> *****************************************************************************************************************
> #I have a series of simulated values, which are means
>
> sim <-
> c(0.0012,0.0009,2,2,9,12,0.0009,2,19,1,1,0.0013,1,0.0009,0.0009,1,26,3,1,2,1,0.0009,1,0.2323,4,2,0.0009,0.0009,0.0009,52,49,1,3,7)
>
> #and actual values
>
> actual <-
> c(0,0,2,0,13,20,0,3,38,0,0,0,1,0,0,0,27,2,0,0,1,0,1,0,4,2,0,0,0,54,21,0,4,11)
>
> #The "X" axes is family, ranging from 1-35, where the "Y" axes is sim and
> actual values.
>
> #What i want to do is plot the simulated values with the 95% CI values, and
> then plot the actual values and see if they fall in the CI'S which they do.
> The idea is that there is no significant difference between the actual
> values and the simulated values.
>
> #I thave Ci for sim and this is where the trouble begins!
>
> simCI <-
> c(0.000908781,0.001248025,0.000928731,0.000885441,0.002384808,0.002700088,0.005377963,0.006202863,0.000918969,0.002566072,0.007687229,0.001593536,0.001578519,0.001299327,0.00217493,0.000908781,0.00090428,0.001550469,0.008840134,0.003300862,0.001546501,0.002775418,0.0014778,0.00090428,0.001546201,0.000898151,0.003446757,0.002854941,0.000863444,0.000918969,0.000924599,0.011732253,0.011488353,0.001788464)
>
> # i then put this in a dataframe
> simvsact <- data.frame(sim = sim, actual = actual, simCI.lower = sim -
> simCI, simCI.upper = sim + simCI, fam = factor(paste('Family', 1:34, sep =
> '')))
>
>
> *****************************************************************************************************************
>
> As afore mentioned i was looking at getting a x/y scatter plot ( i think
> this would be best, if not other suggestions would be greatly appreciated)
> with the CI range block highlighted and the actual line a different colour
> running through the CI range.
>
> I hope this makes sense.
>
> Peter
>
> ______________________________________________
> R-help@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
