R 2.7.2
PPC Mac OS X 10.4.11
library mice 1.13.1
I try to use mice for multivariate data imputation.
My variables are numeric, factors, count data, ordered factors.
First I created a vector for the methods to use with each variable
ImpMethMice<-c(rep("logreg", 62), rep("polyreg",1),
rep("norm",12),
rep("polyreg",12))
next step was
Test<-mice(df, im=ImpMethMice)
I got the following error message:
iter imp variable
1 1 variablenameFehler in impute.logreg(c(2L, 2L, 2L, 2L, 2L, 2L, 2L,
1L, 2L, 2L, 1L, 2L, :
Dimensionen [Produkt 119] passen nicht zur L?nge des Objektes [122]
Zus?tzlich: Es gab 50 oder mehr Warnungen (Anzeige der ersten 50 mit
warnings())
1 1 variablenameError in impute.logreg(c(2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L,
2L, 2L, 1L, 2L, :
dimensions [product 119] don`t match lenght of object [122]
Additional: 50 or more warnings (show first 50 with warnings())
warnings()
1: In any(predictorMatrix[j, ]) ... : wandle Argument des Typs 'double'
nach
boolesch
1: In any(predictorMatrix[j, ]) ... : transform argument of type
'double' to
boolesch
I would be very happy if somebody could help me to fix this.
Thanks in advance.
B.
-----
The art of living is more like wrestling than dancing.
(Marcus Aurelius)
--
View this message in context:
http://www.nabble.com/problems-using-mice%28%29-tp18358093p18358093.html
Sent from the R help mailing list archive at Nabble.com.
