similar to: How might I -remove- a tree from a random forest?

I have a very large data frame, and I'm doing a conversion of all columns into factors. Takes a while (thanks to folks here though, for making faster!), but am wondering about optimization from a memory perspective... Internally, am I better off assigning into a new data frame, or doing one of these: dataframe<-someoperation(dataframe) It would seem that re-assigning into the same data

I've seen how to replace the NA's in a single column with a data frame *> mydata$ncigs[is.na(mydata$ncigs)]<-0 *But this is just one column... I have thousands of columns (!) that I need to do this, and I can't figure out a way, outside of the dreaded loop, do replace all NA's in an entire data frame (all vars) without naming each var separately. Yikes. I'm racking my

Just out of curiosity, I took the default "iris" example in the RF helpfile... but seeing the admonition against using the formula interface for large data sets, I wanted to play around a bit to see how the various options affected the output. Found something interesting I couldn't find documentation for... Just like the example... > set.seed(12) # to be sure I have

"classwt" in the current version of the randomForest package doesn't work too well. (It's what was in version 3.x of the original Fortran code by Breiman and Cutler, not the one in the new Fortran code.) I'd advise against using it. "sampsize" and "strata" can be use in conjunction. If "strata" is not specified, the class labels will be used.

Sorry for the repost, but I've really been looking, and can't find any syntax direction on this issue... Just browsing the documentation, and searching the list came up short... I have some unbalanced data and was wondering if, in a "0" v "1" classification forest, some combo of these options might yield better predictions when the proportion of one class is low (less

You can test if the problem is accumulation in memory registers, which is certainly what this sounds like. Just do a loop over a reasonably small number of iterations and store or print the time between each iteration. If memory accumulation it will run optimally for the first few iterations, after which the time will increase noticeably (essentially exponentially, hence ultimately freezes up). If

Ok, I know R isn't an optimal environment for looping (or so I've heard) but I have a need to loop through columns of data and create new columns of data based on calculations within rows... I'm sure there's a help file, but I'm not sure what search terms to use to find it! The problem is that these new columns need to have names that I can later access... Like NewVar1,

The National Ambulatory and Medical Care Survey is a free data set from the CDC that I'd like to analyze using the "Survey" package in R. Before I dive in, though, it occurred to me that someone may already have gone to the trouble of writing code that will bring in the data and assign the variable names and value labels. This is a big file, so doing it from scratch will take

Hello, all-- I am having some fun playing with the graphing in quantmod-- very nice! I am writing a function to calculate (and hopefully plot) support and resistance lines, but the usual plot call of "abline(h=value)" does not seem to work. Here's my code: require(quantmod) AAPL<-getYahooData("AAPL") candleChart(AAPL,subset="last 3

I have data where where I've taken some measurements three times... twice in rapid succession so I could check test-retest reliability of a piece of equipment, and then a third measurement some time later. Not I'd like to do an analysis where I have two scores... the first being the mean of the first two taken the same day, and the second being the one taken later. I have a lot of

Hi, folks! I'm trying to pull in data using read.csv("my URL goes here"), and it really works fantastically. Amazing to pull in live data right off the internet, into RAM, and get busy... however... occasionally there is a server problem, or the data are not up yet, and instead of pushing through a nice CSV file, the server sends a 404 "Not Found" page... Since the

Just giving the survey package a spin... I'm accustomed to stata, and it seems very similar in many respects. One thing is throwing me, however. I've gotten my data in, and specified the design. Looks like the weighting is right (based on published population estimates from these data), but now I'd like to check my "marginal means" for proportions against those that have

I know I've done this before, but it's been a while and I can't find quite what I need in the help files or archives. I have a text field in a very large data frame. I'd like to add a column that represents the value from an existing field, from the next record (the data are sorted). I'm trying to represent "what happens tomorrow", so the "today" row would

Not sure where to even start on this.... I'm hoping there's some debugging I can do... I have a loop that cycles through several different data sets (same structure, different info), performing randomForest growth and predictions... saving out the predictions for later study... I get about 5 hours in (9%... of the planned iterations.. yikes!) and R just freezes. This happens in

Hello, all. I wanted to use the "survey" package to analyze data from the National Ambulatory Medical Care Survey, and am having some difficulty translating the analysis keywords from one package (Stata) to the other (R). The data were collected using a multistage probability sampling, and there are variables included to identify the sampling units and weights. Documentation from the

Just browsing the documentation, and searching the list came up short... I have some unbalance data and was wondering if, in a "0" v "1" classification forest, if these options might yield better predictions when the proportion of one class is low (less than 10% in a sample of 2,000 observations). Not sure how to specify these terms... from the docs, we have: classwt: Priors

I'm working with the randomForest package and have successfully build a model. I'd like to go one step further however, and use the output from the model to construct a map using the output. My input data are spatial, and I have an independent set of rasterize maps for each of the predictor variables, to map the classification solution. Using the getTree function: >

Hello, I have a question concerning decision trees coming from RWeka : library(RWeka) m =J48(Species~.,data=iris) How could such a decision tree be transferred into a matrix, pretty much in the same fashion, as it is done by getTree() in library(ofw) library(ofw) data(srbct) attach(srbct) ##ofwCART learn.cart.keep <- ofw(srbct,

Hi All, In the script below, the importance measure for column 4 (ie MeanDecreaseGini) indicated "Inf" for V7. Running the getTree command showed that "V7" had been selected at least twice in one of the trees for Random Forest. So the "Inf" command was not generated as a result of dividing the sum of the decreases by 0. Any suggestions on what may be causing the

What does a status value of -3 mean when I do a regression with RF and use the getTree function? left daughter right daughter split var split point status prediction 1 2 3 11 4.721000e+03 -3 15.8489576 2 4 5 5 6.500000e+00 -3 11.3240895 3 6 7 10 6.790674e+02 -3 23.7250079 4