[R] Error with Design.Function(fit)

[Frank E Harrell Jr]

x wrote:
> Hi all,
> 
> I'm reposting this with a more appropriate subject.
> 
> Do I need to define limits as the error message seems to suggest? If so, how? The error message, my code, the output and the first few lines of my data are all below.
> 
> Thank you!
> 
> "Error in Getlim(at, allow.null = TRUE, need.all = TRUE) :
>   variable dmodel.df does not have limits defined in fit or with datadist"
> 
> My code:
> ==================
> library(Hmisc); library(Design); library(lattice); 
> 
> dmodel.df = read.table("./data_cub3.txt", header=TRUE, nrows=100)
> f <- ols(dmodel.df$y1 ~ rcs(dmodel.df$x1,3) )

Why did you ignore the reply from another user earlier today?

library(Design)  # implies library(Hmisc)
df <- read.table(...)
dd <- datadist(df); options(datadist='dd')
f <- ols(y1 ~ rcs(x1,3), data=df)
plot(f, x1=NA, ...)

Frank

> print(f)
> 
> dd <- datadist(dmodel.df$x1)
> options(datadist="dd")
> describe(dmodel.df)
> 
> print( Function(f) )
> plot(dmodel.df$x1, dmodel.df$y1)
> plot(f, add=TRUE, col="blue", pch=2)
> 
> Output:
> ====================
> Linear Regression Model
> 
> ols(formula = dmodel.df$y1 ~ rcs(dmodel.df$x1, 3))
> 
>          n Model L.R.       d.f.         R2      Sigma
>        100      501.6          2     0.9934      45128
> 
> Residuals:
>      Min       1Q   Median       3Q      Max
> -68644.8 -31355.6   -849.9  31823.3 154196.6
> 
> Coefficients:
>            Value Std. Error      t  Pr(>|t|)
> Intercept  35925    12789.6  2.809 0.0060121
> dmodel.df  -1620      422.9 -3.832 0.0002260
> dmodel.df' 25202      523.9 48.102 0.0000000
> 
> Residual standard error: 45130 on 97 degrees of freedom
> Adjusted R-Squared: 0.9932
> 
> Error in Getlim(at, allow.null = TRUE, need.all = TRUE) :
>   variable dmodel.df does not have limits defined in fit or with datadist
> 
> Sample data:
> =========================
> config    benchmark    x1    noise    y1
> 1    verify2    1    0.72    282
> 2    verify2    2    1.6    256
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 


-- 
Frank E Harrell Jr   Professor and Chair           School of Medicine
                      Department of Biostatistics   Vanderbilt University