# setwd("C:/cwru/classes/epbi473_05/lec3(SEER.Abomb)")
# This code convers big SEER text files (with no header or field spacings) into a smaller text file (~33% reduced) with tab spacing and a header row
# This takes about an hour to run, but should only need to be done once for a given SEER dataset  ... make sure all the fields you want are here!

cancers=c('breast','digothr','malegen','femgen','other','respir','colrect','lymyleuk','urinary','test') # test was for debugging
for (pre in c('yr1992_2002.sj_la_rg_ak/','yr1973_2002.seer9/'))
#for (pre in c('yr1992_2002.sj_la_rg_ak/')) # for debugging
for (k in 1:9)
#for (k in 10:10)
{
iname=paste("c:/SEER02/",pre,cancers[k],'.txt',sep="")
oname=paste("c:/SEER02/",pre,cancers[k],'.stx',sep="") 
fid=file(iname,'r')
fid2=file(oname,'w')
y=paste("region","personID","bYear","race","sex","sequen","yDiag","morph","BT","rad","ageRC","stage","NHL","surv","ICD9","COD","numPrim",sep="\t")
writeLines(y,con=fid2)
while (1) {
s=readLines(fid,n=1,ok=T)
#print(s) # for debugging
if (length(s)==0) break
region=substr(s,1,2)
personID=substr(s,3,10)
bYear=substr(s,20,23);
race=substr(s,27,28);
sex=substr(s,30,30);
sequen=substr(s,32,33);
yDiag=substr(s,36,39);
morph=substr(s,45,48);
#bh=substr(s,49,49);
BT=substr(s,50,50);   # 5 = T, 6 = B 
rad=substr(s,67,67);  # 0 or 7 = no radiation, 8,9 = unknown
ageRC=substr(s,84,85);
stage=substr(s,86,86);
NHL=substr(s,89,90);
srvy=substr(s,91,92);
srvm=substr(s,93,94);
ICD9=substr(s,95,98);
COD=substr(s,137,141);
numPrim=substr(s,143,144)
v1=as.numeric(srvy);
v2=as.numeric(srvm);
surv=as.character(12*v1+v2); # months of survival
y=paste(region,personID,bYear,race,sex,sequen,yDiag,morph,BT,rad,ageRC,stage,NHL,surv,ICD9,COD,numPrim,sep="\t")
writeLines(y,con=fid2)
}
close(fid)
close(fid2)
}