pmml - how do I set missing values using PmmlTransformation in r -

i have dataframe in r has na values in it. how can use pmmltransformations set missing value treatment these fields. ive seen can set missingvalue treatments when transforming data (normalization, field mapping, etc) know how set missing values out having normalize data.

    library(pmml)     library(pmmltransformations)      df <- data.frame(id=1:5, y=1:5, x=c(2,4,3,na,8))     databox <- wrapdata(df)      # update wrapped data set x=1 when na      fit <- glm(formula=y~x, data = databox$data)      pmml(fit, transforms=databox) 

many in advance

andrew

if want add missingvaluereplacement=1 attribute miningfield elements in pmml document, append unknownvalue = 1 pmml::pmml.glm function call:

library(pmml) df <- data.frame(id=1:5, y=1:5, x=c(2,4,3,na,8)) # set missing values 1 before training glm model df$x[is.na(df$x)] = 1 fit <- glm(formula=y~x, data = df) # encode information missing value transformation pmml document pmml = pmml.glm(fit, unknownvalue = 1) savexml(pmml, "glm.pmml") 

sure, unknownvalue parameter appears deprecated, need without firing complex sequence of transformations.