se of points, e.g., points(X,Y,pch=15). For each for each x-coord I have multiple y values e.g., (1,2), (1,3), 1,5). How can I plot this data in such a way
ames), then when max is performed, return the name of the column. Then, you can use following function WhereIsMax_function(N, col = T,first=T) { M <- N dim
S+2000 R2, Win98 I searched the archive and did not find the following approach, so I thought I would post it. I _really_ needed greek letters in axis labels. No time to install/learn pstricks, and t
to which Prof. Genton replied: Actually, it is important to Prof. Genton. He just doesn't realize it. If he'd write clean code, more people might assist him with his problem. To put it another way, f
running S-PLUS 2000 on Windows 98 At first I was quite pleased to see that S-PLUS 2000 would import Excel files. My primary databases all export to Excel, anad the conversion therefrom to .csv for re
Some commands permit the user to specifiy a dataframe with conditional operators, e.g., data.frame=fred.dat[ , X>3] in order to operate only on those values for which X >3. I wish to do the same for
I have received the following suggestions: use xyplot() from trellis, which takes a formula and dataset as arguments (Bill Dunlap, Anne York, R.\ Lancelot) plot(y~x,data=fred.dat[fred.dat[,"C"]=="A1"
Let "lik" be the likelihood of an invidual observation. Doing some replication work, the other guy (using an inferior package) encountered a problem with log(lik) when lik is close to zero. Using S-P
Using S-PLUS 2000 Release I on a 300MHz PII with 48M Ram and 100M free on the harddrive. For each of 6000 radio stations I have the following variables: revenue (numeric), market (factor), and owner
S.D. Byers and Julian Taylor provided simple tests for 6000 observations of the form N<- 6000 revenue <- rnorm(N,mean=100, sd=10) market <- factor(sample(1:70, replace=T, size=N)) owner <- factor(sam
James Holtman and Bert Gunter both did a simple multiplication: 270 markets x 1932 owners = matrix with 521,640 cells and I have but 48M RAM. Bert also added something worth repeating: "The moral of
Bill Venables suggested: As another example, I have bootstrapped a few nonlinear models, and I have found that using the solution from the previous iteration can work well. Bruce -- This message was
Not meaning to be contentious, I wish Professor Ripley would please explain why this is "quite rightly so". Ripley (1990) suggests that an RNG, among other things, (1) have a very long period and (2)
Many thanks to Professor Professor Ripley for clarifying his position. For those of you who are unaware, Professor Ripley wrote one of "The Books" realted to the use of RNGs (Stochastic Simulation, W
Indeed, the cycle of the first 16 bits is the period of the sequence. Thanks for pointing out my error. Bruce -- This message was distributed by s-news@wubios.wustl.edu. To unsubscribe send e-mail to
I have found that supplying analytic first derivatives using deriv() can eliminate this problem (not always, but sometimes). Supplying analytic derivatives is well-described in V&R. Bruce -- This me
In replicating some work done with package GAUSS, I wish to maximize a log-likelihood function. Naturally, I want to use analytic derivatives. However, "sneq" (the name given to function) does not ap
I have a vector of radio station revenues whose names are the call sign of the station: KPLA KBXR KFRU KOQL NA NA NA NA 1200 1150 1150 1150 NA NA NA NA I can apply max() to find that the maximum is 1
Many thanks for all the quick solutions. I tested them all, and they all do what I want. I can now continue with my efforts to count the number of angels on the head of a pin. Summary below. Bruce Mc