[R] Error with epiR and sensitivity

Aimee Kopolow alj27 at georgetown.edu
Thu Mar 14 01:55:24 CET 2013


Hi everyone,


I emailed yesterday thinking I had a problem with reading matrix
information as numerical output into epiR. After working on the data
frame today, I realise it's a problem with the data itself rather than
the syntax.


An original data frame of
                             min          max
T1       1.500000e+01 3.999954e+01
SE1      0.000000e+00 1.000000e+00
PRE      0.000000e+00 1.000000e+00
WET      0.000000e+00 5.980000e+00
BE1      4.664642e+00 5.866620e+00
Kappa1   5.500000e+03 2.000000e+04
Kappa3   1.000000e+04 2.000000e+04

produces a latin hypercube sampling set  called samplevalues:


                T1          SE1                       PRE
 WET            BE1            Kappa1   Kappa3
1    31.35590 0.7066388715 0.8665111432 4.965701530 5.783424 12240.019 12675.12
2    28.27640 0.5442730461 0.7000693454 3.181014435 5.183708 16626.566 10759.27
3    28.14695 0.6295741145 0.7818034368 2.262515130 4.670685 16930.360 13857.44
4    30.51873 0.3983581045 0.4026640041 2.730221171 5.058697 19546.625 14408.89
5    16.03162 0.0440886703 0.9954737808 1.002989298 5.310149 13188.279 19500.85
6    19.48413 0.4280443098 0.8500412067 1.668042962 5.068510 11742.748 18891.87
7    36.44783 0.5033961511 0.8249423312 5.582521574 4.722634  8738.121 16457.21
8    39.76318 0.8805976090 0.3430379347 4.876022801 4.787737 19873.134 18660.02
9    39.99782 0.4109272317 0.6606016486 0.191627831 5.625588 11086.803 13569.30

I have produced a column (r1) where each cell is a vector of the
previous cells in that row. I tried binding them with
zz<-cbind(samplevalues, r1) and then running epi<-epi.prcc(zz,
sided.test=2) which is what produced the original error message of

"Error in solve.default(C) :
  Lapack routine dgesv: system is exactly singular"

I then tried 'sensitivity' with the syntax zz<-pcc(samplevalues, r1).
I didn't get any error messages, but print(zz) only reveals the Call
value, and no information.

Any advice on what could be the problem??

thank you,
Aimee Kopolow



More information about the R-help mailing list