similar to: Testing loess fit versus linear fit.

I am trying to (semi) calculate the confidence intervals for a loess smoother (function: loess()), but have been thus far unsuccessful. The CI for the loess predicted values, yhat, are apparently yhat +- t*s * sqrt(w^2), where s is the residual sum of squares and w is the weight function Correct me of I'm wrong, but R uses the tricubic function (1-abs(z)^3)^3, where z = (x-xi)/h, where h

Hi all, I 've got a question about the usage of loess in the modreg package. The documentation (loess.html) states that the smoothing window is either set by span or enp.target. If span is used, the details section of the docs state... <SNIP> DETAILS Fitting is done locally. That is, for the fit at point x, the fit is made using points in a neighbourhood of x, weighted by their

Hi all, I 've got a question about the usage of loess in the modreg package. The documentation (loess.html) states that the smoothing window is either set by span or enp.target. If span is used, the details section of the docs state... <SNIP> DETAILS Fitting is done locally. That is, for the fit at point x, the fit is made using points in a neighbourhood of x, weighted by their

Hello, I was hoping that someone familiar with the implementation details of the loess algorithm might be able to help me resolve some difficulties I am having. I am attempting to reproduce some of the functionality of the loess() function in C++. My primary motivation is that I would like to understand the algorithm in detail. So far I have managed to create a working port in C++ for the

Hi, I wonder why my attempt to extend an existing loess fit to a new data set is producing error. I was trying the following: dat = read.csv(choose.files()) x = dat[,2]; y = dat[,1] x.sort = sort(x) y.loess = loess(y~x, span=0.75) # For testing the above fit with a new dataset: test = read.csv(choose.files()) # test data new_x = test [,1]; new_y = test[,2] new_x.sort = sort(new_x) predicted

Hi, I'm trying to normalize data by fitting a line through the highest density of points (in a 2D plot). In other words, if you visualize the data as a density plot, the fit I'm trying to achieve is the line that goes through the "crest" of the mountain. This is similar yet different to what LOESS does. I've been using loess before, but it does not exactly that as it takes

I'm at my wit's end, and have searched all of my sources. I need to generate a relatively large number of individual LOESS fits each month of data (I have about 16 months of data). Fitting the polynomial is not my problem, figuring out what the formula that describes that polynomial is. apologies in advance if this is so simple, but I need a hand here. Thanks. -- View this message in

Dear all, I have some questions regarding the validity an implementation of statistical tests based on linear models and loess. I've searched the R-help arhives and found several informative threads that related to my questions, but there are still a few issues I'm not clear about. I'd be grateful for guidance. Background and data set: I wish to compare the growth and metabolism

Background: I'm a student of Prof. Cleveland at Purdue University. Eventually, we'd like to release a new version of the loess routine in R. For starters, this implementation would have support for local polynomial degree 3, better control over the number of cells in the KD tree, and perhaps a better solution in higher predictor dimension. I see that Prof. Ripley was responsible for

Greetings and happy new year! I am in the process of converting some of the old S-PLUS scripts from Visualizing Data (Cleveland, 1993) into lattice. In fact, I did most of it several years ago, and at the time, all of the scripts that contained loess() worked fine. Tonight, I ran most of the scripts again, but every one that I tried with a loess() call crashed R. I tried it in two sessions, one

Please do as we ask (repeatedly) and study the help page before posting. 'family' is a separate argument, not part of loess.control, as the help page correctly documents. If you use cars.lo2 <- loess(dist ~ speed, cars, family = "symmetric", control = loess.control(surface = "direct", iterations = 20)) cars.lo2$pars$iterations it prints *20*, as it is

Dear R developers! Considering I got no response or comments in the general r-help forum so far, perhaps my question is actually better suited for this list? I have added some more hopefully relevant technical details to my original post (edited below). Any comments gratefully received! Best regards, David Kreil. ---------- Dear R experts, I have a problem that is a related to the question

Good morning, I am Marta Colombo,student at Politecnico,Milan. I am studying local regression models and I am using loess function. My problem is that when I have a loess object I don't know how to display the fitted surface; in fact, while in S when you have a loess object you can see it writing plot(object), in R this dosen't work. Also I'd like to know if there is something like the

Dear R experts, I have a problem that is a related to the question raised in this earlier post ??? https://stat.ethz.ch/pipermail/r-help/2007-January/124064.html My situation is different in that I have only 2 predictors (coordinates x,y) for local regression but a number of global ("parametric") offsets that I need to consider. Essentially, I have a spatial distortion overlaid over a

Hi, I am segfaulting when using predict.loess() (checked with r62092). I've traced the source with the help of valgrind (output pasted below) and it appears that this is due to int overflow when allocating an int work array in loess_workspace(): liv = 50 + ((int)pow((double)2, (double)D) + 4) * nvmax + 2 * N; where liv is an (global) int. For D=1 (one x variable), this overflows at

Hello, I found a little BUG in loess <stats>. It does not receive the iterations parameter. It can be debugged in the following way: THIS IS AN EXCERPT FROM THE CODE: .... fit <- simpleLoess(y, x, w, span, degree, parametric, drop.square, normalize, control$statistics, control$surface, control$cell, iterations, control$trace.hat) Replace argument iterations with

Dear R experts, I have a problem that is a related to the question raised in this earlier post https://stat.ethz.ch/pipermail/r-help/2007-January/124064.html My situation is different in that I have only 2 predictors (coordinates x,y) for local regression but a number of global ("parametric") offsets that I need to consider. Essentially, I have a spatial distortion overlaid over a

I'm trying to do a Hosmer-Lemeshow 'goodness of fit' test on my logistic regression model. I found some code here: http://sas-and-r.blogspot.com/2010/09/example-87-hosmer-and-lemeshow-goodness.html The R code is above is a little complicated for me but I'm having trouble with my answer: Hosmer-Lemeshow: p=0.6163585 le Cessie and Houwelingen test (Design library): p=0.2843620

I'm updating the loess routines to allow for, among other things, arbitrary local polynomial degree and number of predictors. For now, I've given the updated package its own namespace. The trouble is, panel.loess still calls the original code in package:stats instead of the new loess package, regardless of whether package:loess or package:lattice comes first in the search list. If I

I have a question about running an optimization function on an existing LOESS function defined in R. I have a very large dataset (1 million observations) and have run a LOESS regression. Now, I want to run a Newton-Raphson optimization to determine the point at which the slope change is the greatest. I am relatively new to R and have tried several permutations of the maxNR and nlm functions with