Difference between revisions of "File:Monica.zip"

Give your opinion to the peer rating of the content of this page.
{{ #opasnet_rater: }}

#File: http://www.tilastotiede.fi/ReplicaX/example_MONICA.r
#Author: Juha Karvanen (juha.karvanen at iki.fi)
#
#Summary: This code demonstrates the creation of data replicas with ReplicaX R functions. 
#The R functions needed are available as http://www.tilastotiede.fi/ReplicaX/ReplicaX_functions.r
#For information see http://www.tilastotiede.fi/ReplicaX/ReplicaX_esitys.pdf (in Finnish)
#
#The data used is a published sample of WHO MONICA data. 
#Before running the code, you should download the data from
#http://www.thl.fi/publications/monica/monograph_cd/data/form04_3.zip
#
#The data documentation can be found at
#http://www.thl.fi/publications/monica/monograph_cd/formats/form048.htm
#http://www.thl.fi/publications/monica/monograph_cd/formats/survey.htm
#
#For more information on the MONICA project see http://www.thl.fi/monica/


print("ReplicaX demonstration with the MONICA data. Running the code takes a couple minutes.")

#rm(list=ls())
set.seed(20131103) #to reproduce the results
#path <- "/home/jutakarv/Documents/replica_gen/"
#path <- "/home/q/tiede/replica_gen/"
#path <- paste(getwd(),"/",sep="")
#source(paste(path,"ReplicaX_functions.r",sep=""))

path <- "//cesium/jtue$/_Downloads/form04_3/" # This was the temporary file on my own computer when doing the conversion.
if(!file.exists(paste(path,"form04_3.txt",sep="")))
{  
  stop("Data file form04_3.txt does not exist in the current path. 
  Please download it from http://www.thl.fi/publications/monica/monograph_cd/data/form04_3.zip
  or change the path definition")
}
widths <- c(
	2,1,2,2,6,1,3,8,8,1,1,1,1,2,1,3,1,1,4,1,3,2,1,3,1,3,2,1,1,1,1,1,1,1,1,1,2,1,1,3,3,2,3,3,2,1,2,2,4,2,3,3,8,3,3,8,3,4,2,3,4,
	4,4,2,1,2,1,1,1,1,1,4,4,4,4
)

colnames <- c(
	"form","versn","centre","runit","serial","numsur","samunit","dexam","mbirth","agegr","sex","marit","edlevel",
	"school","cigs","numcigs","daycigs","evercig","stop","iflyear","maxcigs","cigage","cigarsm","cigar","pipesm","pipe",
	"othersm","hibp","drugs","bprecd","hich","chdt","chrx","chrecd","asp","menop","agem","horm","pill","syst1",
	"diast1","rz1","syst2","diast2","rz2","cuff","arm","bpcoder","timebp","rtemp","chol","choldl","dchol","hdl",
	"hdldl","dhdl","scn","cotin","carbmon","height","weight","waist","hip","whcoder","oversion","eage","eageg",
	"cohort1","cohort2","edtert1","edtert2","systm","diastm","chola","hdla"
)

monica <- read.fwf(file=paste(path,"form04_3.txt",sep=""),widths=widths,col.names=colnames)

selectedcolnames <- c("marit","edlevel","school","height","weight","waist","hip")
monica2 <- monica[,selectedcolnames]
vartype <- c("discrete","discrete","discrete","continuous","continuous","continuous","continuous")

#decoding  missing values
monica$height[monica$height==999] <- NA
monica$weight[monica$weight==9999] <- NA
monica$waist[monica$waist==9999] <- NA
monica$hip[monica$hip==9999] <- NA
monica$school[monica$school == 99] <- NA

#unit transformations
monica$weight <- monica$weight/10 #transform to kg
monica$waist <- monica$waist/10 #transform to cm
monica$hip <- monica$hip/10 #transform to cm

write.csv(monica, "//cesium/jtue$/_Documents/Monica.csv")

# This csv file was then zipped and uploaded to [[:File:Monica.zip]]