ArrayStatistics Application User's Guide Version 1.0

Overview

This is an application for computing statistics of data in a simple ASCII array format. Given a collection of similarly formatted data, this application can compute either the mean, the variance, or both. The default format is an ASCII format with space and newline delimiting. Alternatively, the user can specify the format to be UIF/FPIC format. The output format will always match the input format. When specifying the collection of data, the user may pass the filenames as command line arguments, or they may specify a list of filenames.

All computations are done in 64-bit floating-point precision, regardless of the input data precision. There are no constraints on the format; however, all of the input files must have a consistent format. There may be an arbitrary number of rows and columns of data.

Examples

Consider two data files, 'a.txt' and 'b.txt':

$ cat a.txt 
1.2 6.3
6.5 7.8
0.4 4.2

$ cat b.txt 
0.5 1.2
5.4 9.1
3.1 0.9

The mean of these files is computed with the following call:

java -jar ArrayStatistics.jar a.txt b.txt -outputMean mean.txt

which creates the file 'mean.txt':

$ cat mean.txt 
0.85 3.75
5.95 8.45
1.75 2.5500000000000003

Similarly, the variance of the data is computed with the following call:

java -jar ArrayStatistics.jar a.txt b.txt -outputVar var.txt

which creates the file 'var.txt':

$ cat var.txt 
0.12249999999999998 6.5024999999999995
0.3024999999999998 0.4224999999999999
1.8225000000000002 2.7225

If both the mean and variance are needed, it is more efficient to compute them simultaneously with the following call:

java -jar ArrayStatistics.jar a.txt b.txt -outputMean mean.txt -outputVar var.txt

For reference, here is a call that verbosely computes the variance of a set of FPICs that are listed in a file 'dataFilenames.txt', writing the output to 'var.uif':

java -jar ArrayStatistics.jar -verbose -uif -inputFileList dataFilenames.txt -outputVar var.uif