musx

MtGox trades history download

Sep 2nd, 2013
18
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
  1.  
  2. # MtGox trades history download ----------------------------------------
  3.  
  4. #official topic of the script: https://bitcointalk.org/index.php?topic=286755
  5. #caused by mtgox issue related to server overload
  6. #the result set will contain join dataset, bitcoincharts data has fewer attribiutes
  7. #it should handle other currencies
  8.  
  9. #install.packages - RUN ONLY FIRST TIME
  10. #install.packages(c('data.table','RCurl','RJSONIO','RSQLite'))
  11.  
  12. #loading packages
  13. require(data.table) #http://cran.r-project.org/web/packages/data.table/index.html
  14. require(RCurl) #http://cran.r-project.org/web/packages/RCurl/index.html
  15. require(RJSONIO) #http://cran.r-project.org/web/packages/RJSONIO/index.html
  16. require(RSQLite) #http://cran.r-project.org/web/packages/RSQLite/index.html
  17.  
  18. #params
  19. market <- 'mtgoxUSD'
  20. output_chunks <- 1000000
  21. output_format <- c('csv','db') #possible values: 'csv', 'db', c('csv','db') #market is also sqlite table name
  22. csv.sep <- ';'
  23. csv.dec <- ','
  24. ssl.verify <- FALSE #before changing to TRUE put certificate file 'ca-bundle.crt' in the working directory, by default directory of the script: http://curl.haxx.se/docs/caextract.html
  25.  
  26. cat('\n',as.character(Sys.time()),': processing started',sep='')
  27. init_time <- proc.time()[[3]]
  28. #download full history from bitcoincharts
  29. #http://bitcoincharts.com/about/markets-api/
  30. cat('\n',as.character(Sys.time()),': downloading trades csv from bitcoincharts',sep='')
  31. bitcoincharts_dt <- data.table(read.csv(paste0('http://api.bitcoincharts.com/v1/csv/',market,'.csv'), header = FALSE, col.names = c('date','price','amount')))
  32. cat('\n',as.character(Sys.time()),': bitcoincharts trades downloaded',sep='')
  33.  
  34. #download most recent trades until reached time from bitcoincharts
  35. #https://bitbucket.org/nitrous/mtgox-api/#markdown-header-moneytradesfetch
  36. cat('\n',as.character(Sys.time()),': downloading recent trades from MtGox',sep='')
  37. since <- paste0(bitcoincharts_dt[NROW(bitcoincharts_dt),date],'000000')
  38. mtgox_dt <- data.table()
  39. repeat{
  40.   mtgox_dt.batch <- fromJSON(rawToChar(getURLContent(curl = getCurlHandle(),
  41.                                                      url = paste0('https://data.mtgox.com/api/2/BTC',substring(market,6,9),'/money/trades/fetch?since=',since),
  42.                                                      useragent = R.version.string,
  43.                                                      cainfo = 'ca-bundle.crt',
  44.                                                      ssl.verifypeer = ssl.verify,
  45.                                                      ssl.verifyhost = ssl.verify,
  46.                                                      verbose = FALSE,
  47.                                                      binary = TRUE)))
  48.   if(mtgox_dt.batch[['result']] != 'success'){
  49.     cat('\n',as.character(Sys.time()),': mtgox api returns error, retrying',sep='')
  50.     next
  51.   }
  52.   mtgox_dt.batch <- rbindlist(mtgox_dt.batch$data)
  53.   mtgox_dt <- rbindlist(list(mtgox_dt, mtgox_dt.batch))
  54.   if(NROW(mtgox_dt.batch) < 1000) break
  55.   since <- mtgox_dt.batch[NROW(mtgox_dt.batch),tid]
  56. }
  57. cat('\n',as.character(Sys.time()),': MtGox recent trades downloaded',sep='')
  58.  
  59. #join datasets
  60. bitcoincharts_dt <- bitcoincharts_dt[,c('date','price','amount','price_int','amount_int','tid','price_currency','item','trade_type','primary','properties') := list(date,price,amount,as.integer(NA),as.integer(NA),as.numeric(NA),as.character(NA),as.character(NA),as.character(NA),as.character(NA),as.character(NA))]
  61. mtgox_dt <- mtgox_dt[,c('date','price','amount','price_int','amount_int','tid','price_currency','item','trade_type','primary','properties') := list(as.integer(date),as.numeric(price),as.numeric(amount),as.integer(price_int),as.integer(amount_int),as.numeric(tid),price_currency,item,trade_type,primary,properties)]
  62. merged_dt <- rbindlist(list(bitcoincharts_dt[date < last(date),],mtgox_dt))
  63. cat('\n',as.character(Sys.time()),': merged ',NROW(bitcoincharts_dt),' bitcoincharts rows and ',NROW(mtgox_dt),' MtGox rows, current total is ',NROW(merged_dt),' rows',sep='')
  64.  
  65. #output to csv
  66. if('csv' %in% output_format){
  67.   cat('\n',as.character(Sys.time()),': exporting to csv',sep='')
  68.   chunks <- ceiling(NROW(merged_dt)/output_chunks)
  69.   for(i in 1:chunks){
  70.     if(i == 1) write.table(merged_dt[(1+(i-1)*output_chunks):min((i*output_chunks),NROW(merged_dt))], file = paste0('trades.',market,'.csv'), row.names = FALSE, sep = csv.sep, dec = csv.dec, col.names = TRUE)
  71.     else if(i != chunks) write.table(merged_dt[(1+(i-1)*output_chunks):(i*output_chunks)], file = paste0('trades.',market,'.csv'), append = TRUE, row.names = FALSE, sep = csv.sep, dec = csv.dec, col.names = FALSE)
  72.     else if(i == chunks & NROW(merged_dt) > (i*output_chunks)) write.table(merged_dt[(1+(i-1)*output_chunks):(i*output_chunks)], file = paste0('trades.',market,'.csv'), append = TRUE, row.names = FALSE, sep = csv.sep, dec = csv.dec, col.names = FALSE)
  73.   }
  74.   cat('\n',as.character(Sys.time()),': exported csv: ',paste0('trades.',market,'.csv'),sep='')
  75. }
  76.  
  77. #output to sqlite db
  78. if('db' %in% output_format){
  79.   db_conn <- dbConnect(SQLite(), dbname = paste0('trades.db'))
  80.   cat('\n',as.character(Sys.time()),': exporting to db',sep='')
  81.   chunks <- ceiling(NROW(merged_dt)/output_chunks)
  82.   for(i in 1:chunks){
  83.     if(i == 1) dbWriteTable(conn = db_conn, name = market, value = merged_dt[(1+(i-1)*output_chunks):min((i*output_chunks),NROW(merged_dt))], row.names = FALSE)
  84.     else if(i != chunks) dbWriteTable(conn = db_conn, name = market, value = merged_dt[(1+(i-1)*output_chunks):(i*output_chunks)], append = TRUE, row.names = FALSE)
  85.     else if(i == chunks & NROW(merged_dt) > (i*output_chunks)) dbWriteTable(conn = db_conn, name = market, value = merged_dt[(1+(i-1)*output_chunks):(i*output_chunks)], append = TRUE, row.names = FALSE)
  86.   }
  87.   dbDisconnect(db_conn)
  88.   cat('\n',as.character(Sys.time()),': exported sqlite: ',paste0('trades.db'),sep='')
  89. }
  90. cat('\n',as.character(Sys.time()),': processing finished, total time ',proc.time()[[3]] - init_time,'s., bitcoin donation: 1BNBQw2k8u3L9DLnfLFAeHjs5vcEbGnarT',sep='')
Advertisement
Add Comment
Please, Sign In to add comment