Hello everyone!I have a data set with missing observations that I am trying to
impute. I am using MICE and I would like the imputed values to all be positive.
I have two types of variables: prices (P1 to P136) and quantities (Q1 to Q136)
and I also want the range of these two types to be different. Besides these
variables. I am using the squeeze function but I am unable to set it such that
I don't have to write the names of all the variables. Also note that I do not
have all the 136 Ps and Qs because since some were completely missing, I had to
drop those variables. But my first Q is called Q1 and the last one Q136 and
similarly for P. I apologize if I am making a silly little error since I am
still learning R.
This is the code that I have and the error that I am getting:
#####Program starts here#####Read data into Rdf <-
read.table("F:\\testC2S4\\new_C2-S4_head_data.txt", header = TRUE);
#, stringsAsFactors = TRUEdfPZ <- df[ ,1:3]dfDV <- df[
,277:311]dfPtemp <- df[ ,seq(4, 275, by = 2)]dfP <- dfPtemp[, sapply(dfPtemp,
function(x) !all(is.na(x)))]dfQtemp <- df[ ,seq(5, 275, by = 2)]dfQ <-
dfQtemp[, sapply(dfQtemp, function(x) !all(is.na(x)))]remove(dfPtemp, dfQtemp)
NewPrimal <- data.frame(c(dfPZ, dfQ, dfP, dfDV), row.names = NULL, check.rows =
F, check.names = F) remove(dfPZ, dfQ, dfP, dfDV)
library(mice)# setting up the default settings for the imputationini <-
mice(NewPrimal, seed = 52275, imputationMethod = "norm",
maxit=0, pred = quickpred(NewPrimal,
mincor = 0.25, minpuc = 0.95, include =
c("blnNewYearsDay", "blnValentinesDay", "blnStPatricksDay",
"blnEaster", "blnMothersDay", "blnMemorialDay",
"blnFathersDay", "blnIndependenceDay", "blnLaborDay",
"blnThanksgiving", "blnChristmas", "Qtr1", "Qtr2", "Qtr3",
"Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep",
"Oct", "Nov", "Dec"), exclude
= c("ActualWeekEndingDate", "CompanyID", "PriceZoneID", "Date", "Year")),
print=FALSE)
meth <- ini$meth # naming the default method in inipred <- ini$pred # naming
the matrix of default predictors in inipost <- ini$post # naming the default
post processing method in inipost["Q1":"Q136"] <- "imp[[j]][,i] <-
squeeze(imp[[j]][,i],c(0,351000))" #setting the range for
quantitypost["P1":"P136"] <- "imp[[j]][,i] <- squeeze(imp[[j]][,i],c(0,50))"
#setting the range for price
# calling mice to impute given default settings:impNP <- mice(NewPrimal, meth =
meth, maxit=5, pred = pred, post = post)
######My error warning####Error in "Q1":"Q136" : NA/NaN argument
In addition: Warning messages:
1: In post["Q1":"Q136"] <- "imp[[j]][,i] <- squeeze(imp[[j]][,i],c(0,351000))" :
NAs introduced by coercion
2: In post["Q1":"Q136"] <- "imp[[j]][,i] <- squeeze(imp[[j]][,i],c(0,351000))" :
NAs introduced by coercion
Any ideas?
Thank you and have a great weekend!
Rita ________________________________________ "If you think education is
expensive, try ignorance"--Derek Bok
[[alternative HTML version deleted]]
______________________________________________
R-help@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.
