R help - Aug 2013 - Datamatrix in R - extracting data from scatterplot

Say I have a dataframe for plotting scatterplot. The dataframe would be
organized in the following fashion (in CSV format):

name ABC EFG1    32  452    56  67
to, say 200 000 entries

I am going to first do a scatterplot, after which I am going to subset a
portion of the dataset into A using alphahull and export them as XYZ. The
script for doing this:

#plot first plot containing all data
plot(x = X$ABC,
y = X$EFG,
pch=20,)
#subset data using ahull. choose 4 points on the plot
A <- ahull(locator(4, type="p", pch=20), alpha=10000)
#exporting subset
XYZ <- {}for (i in 1:nrow(X)) { if (inahull(A, c(X$ABC[i],X$EFG[i])))
XYZ <- rbind(X,X[i,])}

I am getting the following message if the number of data points in the
subset that I choose is too large:Error in if (p[2] > a + b * p[1]) { :
missing value where TRUE/FALSE needed


Does anyone know what might be causing the problem? Any help is much
appreciated!

	[[alternative HTML version deleted]]

On Aug 15, 2013, at 11:45 AM, BoonFei Tan wrote:
> Say I have a dataframe for plotting scatterplot. The dataframe would be
> organized in the following fashion (in CSV format):
> 
> name ABC EFG1    32  452    56  67
> to, say 200 000 entries
> 
> I am going to first do a scatterplot, after which I am going to subset a
> portion of the dataset into A using alphahull and export them as XYZ. The
> script for doing this:
> 
> #plot first plot containing all data
> plot(x = X$ABC,
> y = X$EFG,
> pch=20,)
What's X and why do you have a comma followed by a right-paren?
> #subset data using ahull. choose 4 points on the plot
> A <- ahull(locator(4, type="p", pch=20), alpha=10000)
It is courteous to put the library call that would load the package that has
`ahull`.
> #exporting subset
> XYZ <- {}for (i in 1:nrow(X)) { if (inahull(A, c(X$ABC[i],X$EFG[i])))
> XYZ <- rbind(X,X[i,])}
Are there some missing linefeeds here?
Did you really want XYZ <- rbind(X,X[i,])} and not XYZ <-
rbind(XYZ,X[i,])} ?

Perhaps you should learn to post in plain text, as suggested in the Posting
Guide.
> 
> I am getting the following message if the number of data points in the
> subset that I choose is too large:Error in if (p[2] > a + b * p[1]) { :
> missing value where TRUE/FALSE needed
> 
> Does anyone know what might be causing the problem? Any help is much
> appreciated!
> 
My vote would be ... missing values.

-- 

David Winsemius
Alameda, CA, USA