[R] plotting lines when data has missing/NA values

[(Ted Harding)]

On 08-Jul-10 19:52:36, Justin Fincher wrote:
> Howdy,
>    I have done many searches and can't seem to find a way around this.
>  I am reading in a .csv file where each row is a dataset and each
> column represents a position.  The values are sparse (there are 2003
> positions but usually only 100-200 values) and the idea is to be able
> to plot each dataset (row) and overlay them in different combinations.
>  What I would like to do is have a plot where each data point is
> connected by a line, but since there are numerous NA values between
> each real data value, I have yet to find out to do this.  Essentially
> I would like each data point to be connected to the next non-NA data
> point.  I also thought about subsetting the relevant data points out,
> but am unsure how to do this while retaining the column numbers so
> they will not be affected spatially.  Any help would be greatly
> appreciated.
> 
> - Fincher

The following small artificial example may do what you seem to want
(if the general idea is right, it should be straightforward to
modify the details to taste):

set.seed(54321)
D <- matrix(rep(NA,60),nrow=4)
D[sample(60,20)] <-runif(20)
print(D,3)
#       [,1] [,2]  [,3]   [,4] [,5]  [,6]  [,7]  [,8]
# [1,] 0.825   NA    NA     NA   NA 0.905    NA    NA
# [2,]    NA   NA    NA 0.0763 0.54    NA    NA    NA
# [3,]    NA   NA 0.130 0.2061   NA 0.622    NA 0.330
# [4,] 0.879   NA    NA     NA   NA 0.937 0.204 0.953
#        [,9] [,10] [,11] [,12] [,13] [,14] [,15]
# [1,] 0.0378 0.170    NA    NA    NA    NA    NA
# [2,]     NA 0.484 0.894 0.230    NA    NA    NA
# [3,]     NA    NA    NA    NA    NA 0.131    NA
# [4,]     NA    NA 0.389 0.859    NA    NA    NA
posns<-(1:15)
cols <-c("red","green","blue","yellow")
M<-max(D,na.rm=TRUE)
ix.NA <- which(!is.na(D[1,]))
plot(posns[ix.NA],D[1,ix.NA],pch="+",xlim=c(0,16),ylim=c(0,M))
lines(posns[ix.NA],D[1,ix.NA])
for(i in (2:nrow(D))){
  ix.NA <- which(!is.na(D[i,]))
  points(posns[ix.NA],D[i,ix.NA],pch="+",ylim=c(0,M))
  lines(posns[ix.NA],D[i,ix.NA],col=cols[(i-1)])
}

Hoping this helps!
Ted.

--------------------------------------------------------------------
E-Mail: (Ted Harding) <Ted.Harding at manchester.ac.uk>
Fax-to-email: +44 (0)870 094 0861
Date: 08-Jul-10                                       Time: 21:34:18
------------------------------ XFMail ------------------------------