[R] intervals {nlme} lower CI greater than upper CI !!!????

[David Winsemius]

On Feb 8, 2011, at 4:14 PM, Federico Bonofiglio wrote:

> Hi folks...
>
> check this out..
>
>> GLU<-lme(gluc~rt*cd4+sex+age+rf+nadir+pharmac+factor(hcv) 
>> +factor(hbs)+
> + haartd+hivdur+factor(arv),
> +  random= ~rt|id, na.action=na.omit)
>
>
>> intervals(GLU)$fixed
>                        lower          est.         upper
> (Intercept)     67.3467070345  7.362307e+01  7.989944e+01
> rt               *0.0148050160*  6.249304e-02  1.101811e-01
> cd4             -0.0032133709 -6.566501e-04  1.900071e-03
> sexM             0.9601467541  3.565606e+00  6.171066e+00
> age              *0.2436472425*  3.420025e-01  4.403578e-01
> rfOTHER         -5.0933264232 -1.198439e+00  2.696449e+00
> rfETEROSESSUALE -2.1085096013  1.974725e+00  6.057960e+00
> rfOMOSESSUALE   -5.8156940466 -1.891870e+00  2.031953e+00
> nadir           -0.0063302516  8.604450e-04  8.051142e-03
> pharmacNRTI     -4.2586487617 -1.252531e+00  1.753586e+00
> pharmacPI       -3.3132784651 -8.829819e-01  1.547315e+00
> factor(hcv)1    -0.4338530808  2.374438e+00  5.182729e+00
> factor(hbs)1    -0.7670414150  3.274923e+00  7.316888e+00
> haartd          -0.0837790540 -4.655330e-02 -9.327549e-03
> hivdur          -0.1351044027  8.417971e-02  3.034638e-01
> factor(arv)1    -6.7807591712 -2.075808e+00  2.629143e+00
> rt:cd4          -0.0001150701 -4.667572e-05  2.171870e-05
> attr(,"label")
> [1] "Fixed effects:"
>
> ...check for instance on row 2 and 5 (bold numbers)....isn't   
> something
> really weird there ?
> lower interval greater than upper !!????

Perhaps you need to review your notes on scientific notation:

 > 1.101811e-01 > 0.0148050160
[1] TRUE
 > 0.2436472425 < 4.403578e-01
[1] TRUE

-- 
David
>
>
> Federico
>
> -- 
> *Little u can do against ignorance,....it will always disarm u:
> is the 2nd principle of thermodinamics made manifest, ...entropy in
> expansion.**....But setting order is the real quest 4  
> truth, ......and the
> mission of a (temporally) wise dude.
> *
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> 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.

David Winsemius, MD
West Hartford, CT