WriteData: write *.data files
In aultsch/DataIO: IO Handling for Files Containing Data

Description Usage Arguments Details Author(s) References See Also Examples

Writes Data, DataDefined, Key and Names in the *.data file format. WriteData is the function for writing *.data files.

The *.data file format enables the user to import data to Microsoft Excel, Mathworks Matlab and allows to inspect and edit data within every text editor. *.data file format is a tab separated csv file with specific header. To prevent failures each column is saved with a CRC-32 checksum. Every *.data file contains a DataDefined line which describes the data in each column. The key column is a unique column of integer values. One value codes exactly one data record, where this single line of data equals one observation. Numbers are stored under Data inside the file strings, using DataDefined = 9,0,1. Strings are saved as Names using DataDefined = 6. DataDefined 2,3 and 8 are used for internal purposes.

1 2	WriteData(FileName, Data, OutDirectory = getwd(), Header =c(), DataDefined, Key, Comment, Names)

`FileName`	string, name of file where the data will be stored
`Data`	`[n,k]` matrix or a `data.frame` containing the data
`OutDirectory`	optional, string, name of directory the data will be saved in, default `getwd()`
`Header`	optional, string vector with the names for the key and data columns, by default 'Key' for `DataDefined`==9, 'Cls' for `DataDefined`==3, 'C1', 'C2', 'C3', 'C4' etc. for `DataDefined`==1 or ==2, must not contain whitespace
`DataDefined`	optional,vector of column type: keys (0=ignore, 1=data(numeric), 2(numeric), 3=class(numeric), 6=Names(character), 8=class(character), 9=key), default is only 1's or 6's with leading 9 for the key column
`Key`	optional, a numeric vector with unique number for each line, by default 1:n, n being the number of rows
`Comment`	optional, array of characters to be written in the first line of the file, it will be marked with '%'
`Names`	optional, list or string matrix with text for each line to be put as last column, usually a short identifier

composition of the *.data file:

all comment lines start with '%', all header lines with '#'

After none or one comment line the header starts with the following information:

1.line: number of rows in Key respective data

2.line: number of columns in file, including Key column

3.line: checksums for each column

4.line: names of columns defined by input variable Names, if not defined see below

After this the key column and the data follows, each column separated by a white space, each row separated by new line.

The length of Key has to be ncol(Data) or ncol(Data[[1]]) respectively. If Key is empty it is replaced by a vector 1:ncol(Data) NOTE: there can be only ONE column with dataDefined == 9 i.e. the column containing the Key

If Names is empty, it is replaced by default names: The default name for a DataDefined==9 column is 'Key'. The default name for a DataDefined==3 column is 'Cls'. The default name for DataDefined==1 or ==2 is 'C1', 'C2', 'C3', 'C4' etc. DataDefined may only contain numbers 0,1,2,3,6,8 and 9. Please use DataDefined == 6 only once. If DataDefined is empty or wrong it is replaced by c(9,1,..).

Tim Schneider, Michael Thrun

www.uni-marburg.de/fb12/datenbionik mthrun@mathematik.uni-marburg.de

ReadData

#create a matrix
#mat = matrix( 
	#c(1,0,0,0,1,0,0,0,1), 
	#nrow=3, 
	#ncol=3) 
	
#Write this matrix into the current directory
#WriteData(FileName = "test.data",Data = mat)
	
#Write this matrix into a different directory, for example under MacOS X
#WriteData(FileName = "test.data",Data = mat, InDirectory = '~/Desktop/')