[R] Reading numeric data in Excel as character data in R using RODBC

[White, Charles E WRAIR-Wash DC]

The Excel 'database' I am pulling from mixes multiple character
indicators for different kinds of non-numeric results in the same column
as the numeric values. A simple sqlFetch turns all of the character
indicators into NAs. I would like to pull in all of the Excel data as
character so that I can split the column into multiple types of vectors
and control the type conversion within R. Can anyone provide clues as to
how I go about getting RODBC to bring in data containing some numbers
without running a numeric conversion?

The simple code I use to pull data directly from Excel is generally
something like the following:
prj<-'C:/data' 
library(RODBC)
channel <-
odbcConnectExcel(paste(prj,"/intermediate/datafile.xls",sep=''))
r <- sqlTables(channel)
ifn.display <- sqlFetch(channel, "SHEET1")
il2.display <- sqlFetch(channel, "SHEET2")
pl.display <- sqlFetch(channel, "SHEET3")
odbcClose(channel)

Background: The current problem is small enough to solve with read.csv
but I expect to get larger problems in the future. My clients generally
give me data in Excel format. Sometimes I get multiple character
indicators for different kinds of non-numeric results. Sometimes I get
40+ sheets in a single file where all of the sheets are formatted alike.
I'd like to be ready to run a reasonably automated data conversion
whenever a client decides to give me both at once.

Thanks for your time.

Chuck