Created
November 9, 2010 19:36
-
-
Save ajdamico/669666 to your computer and use it in GitHub Desktop.
download all publicly-available data files for every year of the Medical Expenditure Panel Survey
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| library(RCurl) | |
| setwd("R:\\Medical Expenditure Panel Survey\\Data") | |
| #input all available MEPS public use file numbers | |
| year <- c(1996:2008) | |
| consolidated <- c(12,20,28,38,50,60,70,79,89,97,105,113,NA) | |
| conditions <- c("06r",18,27,37,52,61,69,78,87,96,104,112,NA) | |
| jobs <- c("07",19,25,32,40,56,63,74,83,91,100,108,116) | |
| prpf <- c(24,47,47,47,47,57,66,76,88,95,103,111,119) | |
| longitudinal <- c(23,35,48,58,65,71,80,86,98,106,114,NA,NA) | |
| events <- c(10,16,26,33,51,59,67,77,85,94,102,110,118) | |
| mepsmatrix <- data.frame(year, consolidated, conditions, jobs, prpf, longitudinal, events) | |
| #assign all events files | |
| mepsmatrix$rx <- paste(mepsmatrix$events,"a",sep="") | |
| mepsmatrix$dental <- paste(mepsmatrix$events,"b",sep="") | |
| mepsmatrix$other <- paste(mepsmatrix$events,"c",sep="") | |
| mepsmatrix$inpatient <- paste(mepsmatrix$events,"d",sep="") | |
| mepsmatrix$er <- paste(mepsmatrix$events,"e",sep="") | |
| mepsmatrix$outpatient <- paste(mepsmatrix$events,"f",sep="") | |
| mepsmatrix$office <- paste(mepsmatrix$events,"g",sep="") | |
| mepsmatrix$hh <- paste(mepsmatrix$events,"h",sep="") | |
| mepsmatrix$events <- NULL | |
| ##download special files | |
| download.file( "http://www.meps.ahrq.gov/mepsweb/data_files/pufs/h36u07ssp.zip" , "temp" ) | |
| zipcontents <- unzip( "temp" , list=T) | |
| unzip( "temp" ) | |
| file.rename( as.character(zipcontents[1,1]) , "linkage.ssp" ) | |
| file.remove("temp") | |
| download.file( "http://www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h36/h36u07cb.pdf" , "linkage - cb.pdf" , mode="wb" , cacheOK=F , method="internal" ) | |
| download.file( "http://www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h36/h36u07doc.pdf" , "linkage - doc.pdf" , mode="wb" , cacheOK=F , method="internal" ) | |
| download.file( "http://www.meps.ahrq.gov/mepsweb/data_files/pufs/h36b07ssp.zip" , "temp" ) | |
| zipcontents <- unzip( "temp" , list=T) | |
| unzip( "temp" ) | |
| file.rename( as.character(zipcontents[1,1]) , "linkage - brr.ssp" ) | |
| file.remove("temp") | |
| download.file( "http://www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h36brr/h36b07cb.pdf" , "linkage - brr cb.pdf" , mode="wb" , cacheOK=F , method="internal" ) | |
| download.file( "http://www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h36brr/h36b07doc.pdf" , "linkage - brr doc.pdf" , mode="wb" , cacheOK=F , method="internal" ) | |
| #begin downloading all files for all years | |
| for (i in 1:nrow(mepsmatrix) ) { | |
| for (j in 2:ncol(mepsmatrix) ) { | |
| if (!is.na(mepsmatrix[i,j])){ | |
| #download the SSP file | |
| filenumbername <- paste("h",mepsmatrix[i,j],"ssp.zip",sep="") | |
| meps_zipped_file <- paste("http://www.meps.ahrq.gov/mepsweb/data_files/pufs/h" , mepsmatrix[i,j] , "ssp.zip" , sep="") | |
| err <- try( getURLContent(meps_zipped_file) , silent=T ) | |
| if( class(err) == "try-error" ){ | |
| download.file( sub("ssp.zip","f1ssp.zip",meps_zipped_file) , sub("ssp.zip","f1ssp.zip",filenumbername) ) | |
| zipcontents <- unzip( sub("ssp.zip","f1ssp.zip",filenumbername) , list=T) | |
| unzip( sub("ssp.zip","f1ssp.zip",filenumbername) ) | |
| finalname <- paste( mepsmatrix[i,1] , " - " , names(mepsmatrix)[j] , " f1.ssp" , sep="" ) | |
| file.rename( as.character(zipcontents[1,1]) , finalname ) | |
| file.remove(sub("ssp.zip","f1ssp.zip",filenumbername)) | |
| download.file( sub("ssp.zip","f1ssp.zip",meps_zipped_file) , sub("ssp.zip","f2ssp.zip",filenumbername) ) | |
| zipcontents <- unzip( sub("ssp.zip","f2ssp.zip",filenumbername) , list=T) | |
| unzip( sub("ssp.zip","f2ssp.zip",filenumbername) ) | |
| finalname <- paste( mepsmatrix[i,1] , " - " , names(mepsmatrix)[j] , " f2.ssp" , sep="" ) | |
| file.rename( as.character(zipcontents[1,1]) , finalname ) | |
| file.remove(sub("ssp.zip","f2ssp.zip",filenumbername)) | |
| } else { | |
| download.file( meps_zipped_file , filenumbername ) | |
| zipcontents <- unzip( filenumbername , list=T) | |
| unzip( filenumbername ) | |
| finalname <- paste( mepsmatrix[i,1] , " - " , names(mepsmatrix)[j] , ".ssp" , sep="" ) | |
| file.rename( as.character(zipcontents[1,1]) , finalname ) | |
| file.remove(filenumbername) | |
| } | |
| err <- NULL | |
| #download the codebook if possible | |
| cbname <- paste( mepsmatrix[i,1] , " - " , names(mepsmatrix)[j] , " cb.pdf" , sep="") | |
| cbsite <- paste( "http://www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h",mepsmatrix[i,j],"/h",mepsmatrix[i,j],"cb.pdf",sep="") | |
| err <- try( getURLContent( cbsite ) ) | |
| if (! class(err) == "try-error" ) download.file( cbsite , cbname , mode="wb" , cacheOK=F , method="internal" ) | |
| err <- NULL | |
| #download the documentation if possible | |
| docname <- paste( mepsmatrix[i,1] , " - " , names(mepsmatrix)[j] , " doc.pdf" , sep="") | |
| docsite <- paste( "http://www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h",mepsmatrix[i,j],"/h",mepsmatrix[i,j],"doc.pdf",sep="") | |
| err <- try( getURLContent( docsite ) ) | |
| if (! class(err) == "try-error" ) download.file( docsite , docname , mode="wb" , cacheOK=F , method="internal" ) | |
| err <- NULL | |
| } | |
| } | |
| } | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment