similar to: Extreme AIC in glm(), perfect separation, svm() tuning

Dear R-users, I use glm() to do logistic regression and use stepAIC() to do stepwise model selection. The common AIC value comes out is about 100, a good fit is as low as around 70. But for some model, the AIC went to extreme values like 1000. When I check the P-values, All the independent variables (about 30 of them) included in the equation are very significant, which is impossible, because we

I''m running brglm with binomial loguistic regression. The perhaps multicollinearity-related feature(s) are: (1) the k IVs are all binary categorical, coded as 0 or 1; (2) each row of the IVs contains exactly C (< k) 1''s; (3) k IVs, there are n * k unique rows; (4) when brglm is run, at least 1 IV is reported as involving a singularity. I''ve tried recoding the n

Dear all, I am encountering some issues with my data and need some help. I am trying to run glm analysis with a presence/absence variable as response variable and several explanatory variable (time, location, presence/absence data, abundance data). First I tried to use the glm() function, however I was having 2 warnings concerning glm.fit () : # 1: glm.fit: algorithm did not converge # 2:

Dear List, Apologies for this off-topic post but it is R-related in the sense that I am trying to understand what R is telling me with the data to hand. ROC curves have recently been used to determine a dissimilarity threshold for identifying whether two samples are from the same "type" or not. Given the bashing that ROC curves get whenever anyone asks about them on this list (and

Hi all, I have run a (glm) analysis where the dependent variable is the gender (family=binomial) and the predictors are percentages. I get a warning saying "fitted probabilities numerically 0 or 1 occurred" that is indicating that quasi-separation or separation is occurring. This makes sense given that one of these predictors have a very influential effect that is depending on a

Some of the run commandlines don't seem to be the *default ref* configuration. For instance, in 099.go, what's in the script is: LEVEL = ../../../.. BM=099.go ifeq ($(RUN_TYPE),test) STDIN_FILENAME = /dev/null STDOUT_FILENAME = null.out RUN_OPTIONS = 40 19 else STDIN_FILENAME = 2stone9.in STDOUT_FILENAME = 2stone9.out RUN_OPTIONS = 50 9 endif include ../../Makefile.spec95

I'm trying to sort a DATAFRAME by a column "ID" that contains alphanumeric data. Specifically,"ID" contains integers all preceeded by the character "g" as in: g1, g6, g3, g19, g100, g2, g39 I am using the following code: DATAFRAME=DATAFRAME[order(DATAFRAME1$ID),] and was hoping it would sort the dataframe by ID in the following manner g1, g2, g3, g6, g19,

I'm trying to impute data using the mi package, but after running through almost the entire first round of imputations (which takes quite a while), it throws this error (I'll include the whole output prior to the error for context). Does anyone know what is causing it, or how I can fix it? More specifically, how can I tell what is throwing the error so I know what to fix? Is

On Thu, 21 Jul 2005, Long Fei wrote: > which suggests the compiling command: > gcc -DSPEC_CPU95 -O2 -o spec_go g2.c g22.c g23.c g25.c g26.c g27a.c g27b.c > g28.c g29.c g2eye.c g2jlib2.c g2jos.c g2list.c g2reas.c g2s2.c g2s3.c g2shp.c > and the running commandline: > ./spec_go 50 9 2stone9.in > < are these correct ? > I believe that we only include test/train inputs for

Hello List, I'm trying to help a family from the Dominican Republic and to do so need a temporary DID from DR. The short story is that there is a 2 year old here with a serious heart defect from a remote area of Dominican Republic near the Haitian border. He was referred to Gift of Life D.R. who eventually contacted Gift of Life Central Florida and this is where I'm involved. The child

hi all, I have a binary data set and am now confronted with a "separation" issue. I have two predictors, mood (neutral and sad) and game type (fair and non-fair). By "separation", I mean that in the non-fair game, whereas 20% (4/20) of sad-mood participants presented a positive response (coded as 1) in the non-fair game, none of neutral-mood participants did so (0/20). Thus,

Hi all, As you can see from below, the result is strange... I would imagined that the bb result should be much higher and close to 1, any way to improve the fit? Any other classification methods? Thank you! data=data.frame(y=rep(c(0, 1), times=100), x=1:200) aa=glm(y~x, data=data, family=binomial(link="logit")) newdata=data.frame(x=6, y=100) bb=predict(aa, newdata=newdata,

On Thu, 21 Jul 2005, Long Fei wrote: > Does anyone know the "standard" commandline to compile and run spec95 > benchmarks ? > > I found some scripts in llvm test suite (External/SPEC/CINT95), but I am not > sure if they are the *default ref* configurations. <Some of them seem to use > smaller problem sizes.> SPEC95 was retired many years ago, so few people

Does anyone know the "standard" commandline to compile and run spec95 benchmarks ? I found some scripts in llvm test suite (External/SPEC/CINT95), but I am not sure if they are the *default ref* configurations. <Some of them seem to use smaller problem sizes.> SPEC95 was retired many years ago, so few people around me know how to compile and run them. I am building from

CRAN (and crantastic) updates this week New packages ------------ * allan (1.0) Alan Lee http://crantastic.org/packages/allan Automates Large Linear Analysis Model Fitting * andrews (1.0) Jaroslav Myslivec http://crantastic.org/packages/andrews Andrews curves for visualization of multidimensional data * anesrake (0.3) Josh Pasek http://crantastic.org/packages/anesrake This

Hi, I run the following tuning function for svm. It's very strange that every time i run this function, the best.parameters give different values. [A] >svm.tune <- tune(svm, train.x, train.y, validation.x=train.x, validation.y=train.y, ranges = list(gamma = 2^(-1:2), cost = 2^(-3:2))) # where train.x and train.y are matrix