Re: [R] Quantile regression with complex survey data

edian is 0.50 quantile regression). Bassett et al. 2002. Quantile models and estimators for data analysis. Metrika 55: 17-26. (describes weighted QR for survey of school characteristics and student achievement scores). Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science C

Re: [R] weighted.median function from package R.basic

to estimate medians (minimizes and objective function) and thus might require less computing time on a large data set. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: brian_c...@usgs.gov tel: 970 226

Re: [R] opposite estimates from zeroinfl() and hurdle()

tic regression. I'm guessing that the hurdle model as a two-stage model is using a logistic regression predicting the probability of 1, hence the reversed signs of the estimates in the logistic regression portion of the model. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins

Re: [R] Goodness of fit in quantile regression

local regions of the predictor space, then you might want to check out Stef van Buuren's work on worm plots to diagnose fit in quantile regression. Don't remember where he published this but his email is stef.vanbuu...@tno.nl Brian Brian S. Cade, PhD U. S. Geological Survey Fort

Re: [R] Quantile GAM?

to specifiy the degree of polynomial and number and location of knots. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: brian_c...@usgs.gov tel: 970 226-9326 From: Jonathan Greenberg To: "

Re: [R] Kolmogorov-Smirnov test and the plot of max distance between two ecdf curves

the plot.rqs() function. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: brian_c...@usgs.gov tel: 970 226-9326 From: user1234 To: r-help@r-project.org Date: 10/05/2012 06:46 AM Subject: Re: [R

[R] quantreg installation and conflicts with R 2.15.2

rtup - Warning message: unable to restore saved data in C:\CADESTUFF\DATA\BarryNoon\.RData > Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: brian_c...

[R] Fw: quantreg installation and conflicts with R 2.15.2

Just noticed that I get a similar error about object 'kronecker' in "Matrix" package when trying to load "lme4". So this is a more pervasive problem. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort

Re: [R] lme: subject-specific slopes.

and, thus, in the scale you are expecting. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: brian_c...@usgs.gov tel: 970 226-9326 From: "John Sorkin" To: , "Kenneth Frost" Date:

[R] standard error for predicted mean count from ZIP

Does anyone have code for computing the standard error of the predicted mean count from a zero-inflated Poisson regression model estimated by the zeroinfl() function from the pscl package (and yes, we've checked with A. Z. already)? Thank you Brian Brian S. Cade, PhD U. S. Geological S

Re: [R] fisher exact for > 2x2 table

re statistic will be reasonable and is irrelevant if you calculate your probabilities by exact enumeration of all permutations. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: brian_c...@usgs.gov

Re: [R] Comparing non nested models with correlation coefficients (inspired from Lorch and Myers )

As a follow-up to Greg's suggested graphical presentation, it seems like the Vuong test is sometimes used to compare fits of non nested models. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: br

Re: [R] Higher log-likelihood in null vs. fitted model

sums of squares for ls regression). Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: brian_c...@usgs.gov tel: 970 226-9326 From: Andrew Miles To: Mark Leeds Cc: r-help@r-project.org Date: 05/31/2012

Re: [R] quantile regression: out of memory error

Using tau = -1 is causing rq() to try and estimate all possible quantiles and store the results. With 11253 observations this would be a formidable feat. Try estimating the model with say tau = 1:99/100 to give a more tractable number of estimates. Brian Brian S. Cade, PhD U. S

Re: [R] Quantreg-rq crashing trouble

estimate. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: brian_c...@usgs.gov tel: 970 226-9326 From: Steven R Corsi To: r-help@r-project.org Date: 07/21/2011 04:04 PM Subject: [R] Quantreg-rq

Re: [R] quantreg Wald-Test

citations with details on the test statistics). So this comparison corresponds to the typical linear model hypothesis of a set of parameters equaling zero. I prefer the rankscore test if sample sizes are not too large (e.g., <5,000). Brian Brian S. Cade, PhD U. S. Geological Survey F

Re: [R] Imputing data below detection limit

, etc). Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: brian_c...@usgs.gov tel: 970 226-9326 From: Bert Gunter To: Jessica Streicher Cc: r-help@r-project.org Date: 08/13/2012 09:28 AM Subject: Re

Re: [R] Function to compute multi-response, multi-rater kappa?

P ( we're in the process of porting it to an R package but it will be several months before its ready to go). There are versions of MRPP (less complete) available in the vegan package for R, but I don't know whether it will do the randomized block variant required for Cohen's

[R] weights in lmer

onal bouts on wild horse behavior. So is weights not really yet implemented or am I missing something? m2.wt<- lmer(HD_RATE ~ NUM_PAIR + (NUM_PAIR|HMA),data=out2.5,weights=HOURS/max(HOURS), contrasts=list(NUM_PAIR="contr.treatment"),family=gaussian) Brian Brian S. Cade, PhD

Re: [R] categorical data analysis - fisher.exact for 2x2 and greater

2007. Springer) indicates that these algorithms were only efficient for <= 5 conditional loops so 2x2, 3x2, 4x2, 5x2, 6x2, 3x3, and 2x2x2 were feasible back in the 1990's. I'm not sure whether more recent computing capability permits larger tables now. Brian Brian S. Cade U. S. Geol

Re: [R] Inconsistent linear model calculations

Well at a minimum V4 and V6 (used in lm()) are not the same in these two data sets. Not sure why but there it is. Brian Brian S. Cade U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: [EMAIL PROTECTED] tel: 970 226-9326 e

[R] graphics - line resolution/pixelation going from R to windows metafile

e might be a solution but I've yet to find it. Any suggestions would be welcome. Brian Brian S. Cade U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: [EMAIL PROTECTED] tel: 970 226-9326 [[alternative HTML version

Re: [R] weighted quantiles

ted quantile estimates. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: [EMAIL PROTECTED] tel: 970 226-9326 sj <[EMAIL PROTECTED]> Sent by: [EMAIL PROTECTED] 10/07/2008 12:38 PM To r-hel