similar to: Coefficient of determination -- should be "compare to a trivial model"

Hi All, Can you tell me why I get the error message below? It's driving me nuts. Thanks, Tom > r_squared [[1]] [1] 0.9083936 [[2]] [1] 0.8871647 [[3]] [1] 0.8193883 [[4]] [1] 0.728157 [[5]] [1] 0.8849525 [[6]] [1] 0.8459416 [[7]] [1] 0.6702318 [[8]] [1] 0.02997816 [[9]] [1] 0.8974268 [[10]] [1] 0.881217 [[11]] [1] 0.8006688 [[12]] [1] 0.7207697 [[13]] [1] 0.8703734 [[14]]

Antonio, For linear regression we have the following identity total SS = regression SS + residual SS (*) where total SS is the sum of squares of observations around their mean, i.e. total SS = (n-1)*var(y) and residual SS is given by the deviance function. R-squared is defined as R^2 = regression SS/ total SS = 1 - residual SS/total SS. You can use this last formula

All: I have a SQlite database where I have stored some verification data by date & time (cycle Z/UTC), lead_time as well as type, duration, etc. I would like to analyze & plot the data as monthly averages. I have looked at a bunch of examples which use some combination of zoo and aggregate, but I have not been able to successfully apply bits and pieces from the examples I have found. Any

Ravi Varadhan and I have been looking at UCMINF to try to identify why it gives occasional (but not reproducible) errors, seemingly on Windows only. There is some suspicion that its use of DBLEPR for finessing the Fortran WRITE() statements may be to blame. While I can find DBLEPR in Venables and Ripley, it doesn't get much mention after about 2000 in the archives, though it is in the R FAQ

Ravi Varadhan and I have been looking at UCMINF to try to identify why it gives occasional (but not reproducible) errors, seemingly on Windows only. There is some suspicion that its use of DBLEPR for finessing the Fortran WRITE() statements may be to blame. While I can find DBLEPR in Venables and Ripley, it doesn't get much mention after about 2000 in the archives, though it is in the R FAQ

How about allowing underscore? (I believe WRE is silent on this, and I have not tried submitting a package with underscore in the title.) As I pointed out in my OP, _optim()_ works. And we have the advantage that we can distinguish package from function. The purpose of consistent editing is surely to provide the affordances that save us from needing extra documentation, as per Donald Norman's

I am curious as to the "lm" calculation of R2 (multiple coefficient of determination, I assume) when intercept is zero. I have 18 data points, two independent variables: First, a model with an intercept: > mod0=lm(Div~Rain+Evap,data=test) > summary(mod0)$r.squared [1] 0.6257541 > cor(predict(mod0),test$Div)^2 [1] 0.6257541 The $r.squared and the result from "cor"

Dear R-users, I used lm() to fit a standard linear regression model to a given data set, which led to a coefficient of determination (R^2) of about 0.96. After checking the residuals I realized that they follow an autoregressive process (AR) of order 1 (and therefore contradicting the i.i.d. assumption of the regression model). I then used gls() [library nlme] to fit a linear

Dear all, Is there a instruction that can help me obtain the coefficient of determination R^2 after doing linear/nonlinear regression using lm/nls? [[alternative HTML version deleted]]

Hello everyone, I have estimated system of three linear equations with one non-linear restrictions with nlsystemfit. I was wondering how I can calculate the R-squared (or some alternative coefficient of determination) for the whole system. This is automatically given by linear systemfit but not by nlsystemfit. I can get the values for each of the equations separately, but apparently not for

Hello - I am currently working through some MLR models and wanted to perform cross validation to assess the model's prediction error. As a Minitab user prior to discovering R, I was able to view the PRESS statistic and the resultant cross validation R^2. Is there a canned function already present that does this or should I go ahead and just right a user defined function? Along the same

> On 25 Apr 2015, at 13:11 , Prof J C Nash (U30A) <nashjc at uottawa.ca> wrote: > > Hendrik pointed out it was the parentheses that gave the complaint. > Single quotes and no parentheses seem to satisfy R CMD check. Perhaps > that needs to be in the WRE. Well, it is in ?toTitleCase: ...However, unknown technical terms will be capitalized unless they are single

Dear, I want to compute coefficient of determination (R-squared) to complement AIC for model selection of multivariable GLM. However, I found this is not a built-in function in glm. neither is it available through reviewing the question in the R-help archive. Please kindly help and thanks a lot. Elaine [[alternative HTML version deleted]]

If I'm not mistaken, partial R-squared is the R^2 of the quantities plotted in a partial residual plot, so you can base the computation on that. Prof. Fox's `car' package on CRAN has a function for creating those plots, but you need to figure out the way to extract the quantities being plotted. [In any case, the basic tools for doing such computations are all in R, and it

Dear all, I'm modeling growth curve of some ecosystems with respect to their rainfall-productivity relationship using a simple linear regression (ANPP(t)=a+b*Rain(t)) and a modified version of the Brody Model ANPP(t)=a*(1-exp(-b*rain(t))) I would like to know why the "best model" is function of the criteria that I use (maximizing the fit using R2 or testing the Null hypothesis with

I was preparing a fix for a minor glitch in my optimx package and R CMD check gave an error that the title was not in title case. It is A Replacement and Extension of the optim() Function R CMD check suggests the incorrect form A Replacement and Extension of the Optim() Function 'Writing R Extensions' suggests single quotes, i.e., A Replacement and Extension of the 'optim()'

I've been creating some R tools that manipulate objective functions for optimization. In so doing, I create a character string with R code, and then want to have it in my workspace. Currently -- and this works fine -- I write the code out, then use source() to bring it in again. Example: cstr<-"jack<-function(x){\n cat(\"Silly x:\")\n print(x) \n }\n" write(cstr,

I spent more time than I should have debugging a script because I wanted x<-seq(0,100)*0.1 but typed x<-seq(O:100)*0.1 seq(0:100) yields 1 to 101, Clearly my own brain to fingers fumble, but possibly one others may want to avoid it. JN

Having looked at documentation and codes, I'm still looking to find out who did the installation of the Port libraries. As I work on optimization code improvements, there are often choices of settings (tolerances etc.), and I'd prefer to learn if there are particular reasons for choices before diving in and making any adjustments. It is also worth giving credit where it is due. Given

In building a function for a package, I'd like to set the defaults in a control list, e.g., makeg<-function(parameters, eps = 1.0e-7, control=list(showwork=TRUE, rubbish=1.0e+101)){ etc. } This does not provide showwork or rubbish within the function if control() is not fully specified. Copying others, I've previously set the control defaults within the function then matched