ArtificialData: create artificial data for testing

In MultiJoin: Enables Efficient Joining of Data File on Common Fields using the Unix Utility Join

Description

This function allows quick generation of a test data set which can be used with the majority of the Join functions

Usage

ArtificialData(fakeDataDir = "~/fakeData2/", joinKey = letters[1:20], 


    numFiles = 4, N = rep(15, numFiles), SORT = 1, GZIP = 0, 


    sep = c(" ", ",", "\t", "|")[1], prefix = "file", suffix = ".txt", 


    daten = month.abb, NCOL = rep(3, numFiles), chunkSize = 1000, 


    verbose = 0)

Arguments

fakeDataDir	directory to put the data
joinKey	set of join keys to choose from (has to be longer than N) - this column will be the key for join
numFiles	number of files to split the data across
N	number of rows in each file created, e.g. N = c(15,20,10,30)
SORT	should the join key be sorted?
GZIP	should the data files created by gzipped?
sep	column delimiter; default white space
prefix	file name prefix
suffix	file name suffix
daten	data to sample from
NCOL	number of data columns per file
chunkSize	write that many lines to the file at once
verbose	level of verbosity

Value

invisibly return data and file names

Author(s)

"Markus Loecher, Berlin School of Economics and Law (BSEL)" <[e-mail]>

Examples

if (0){


  ArtificialData("fakeData2",verbose=1)


  ArtificialData("fakeData2",joinKey = 1:2000, N = rep(1500,4) ,verbose=0)


  


  ret = ArtificialData(fakeDataDir="/tmp/fakeData")


  ret = ArtificialData(fakeDataDir="./fakeData", joinKey=letters[1:10], numFiles = 6, N = rep(5,6))


  ret = ArtificialData(SORT = 1, GZIP = 1)





  ret = ArtificialData(fakeDataDir="fakeData", joinKey = 0:9, N = rep(6, 4), verbose=1)


  #on allegro:


  ret = ArtificialData(fakeDataDir="./fakeData", joinKey=letters, numFiles = 10, 


                       N = rep(18,10), NCOL=rep(5,10))


}

MultiJoin documentation built on May 1, 2019, 7:32 p.m.