VERSION sprint 1.0.6

	o This version of SPRINT now works with R 3.0.x.
	
	o Fixed an installation warning that caused the message "warning: argument type 'enum commandCodes *' 
	doesn't match specified 'MPI' type tag that requires 'int *'    [-Wtype-safety]".
	
	o Fixed an intermittent bug in pcor that caused the following warning: "Error in ff(dim = 
	 c(number_of_samples, number_of_samples), dimnames = dimnames_,  :   length exceeds file length".
	 
	o The randomForest library is loaded by default when SPRINT is loaded.
	
	o The Makefile was changed to remove gmake and gcc dependencies.
	
	
VERSION sprint 1.0.5

	o Added pstringdistmatrix(), a function which calculates the hamming distance between strings.
	
	o SPRINT can now be installed with the latest version of mpich which is MPI 3 compliant. 
	
	o The ppam() and pcor() functions now keep the dimnames of the data.
	
	o The dependencies have been updated, so now the call 'install.packages("sprint", dependencies = T)' should work.
	
	o The papply() documentation has been updated to show that it does not currently support ff data types.
	
	o A bug that caused the install error 'C++ MPI compiler not detected' was fixed
	
	
VERSION sprint 1.0.4

	o Added support for Mac OSX and fixed a bug in ppam.

VERSION sprint 1.0.2

    o Fixed the issue with filling the $clusinfo[] and $silinfo$avg.width values in ppam
    o The "bug in C cstat(): ntt=0 !!!" no longer appear in ppam
    o psvm(), a svm function

VERSION sprint 1.0.1

    o Fixed the issue with worker processes finalization.

VERSION sprint 1.0.0

    o First 'full' release of SPRINT.
    
    o Added papply(), a parallel 'apply' function.
    
    o Added pboot(), a parallel bootstrapping function.
    
    o Added prandomForest(), an ensemble tree classifier.
    
    o Added pRP(), a parallel rank product function.

VERSION sprint beta 0.3.0

    o Added ppam(), a parallel clustering function performing a partition
    around medoid.
    
    o Added functionality to pcor(), to allow the correlation between two 
    matrices.

VERSION sprint beta 0.2.0

    o Simplified installation process and reduced set of requirements.
    
    o Added pmaxt(), a parallel permutation test function.
    
    o Added functionality to pcor() to allow the output of the distance matrix 
    instead of the correlation coefficient matrix.

VERSION sprint beta 0.1.0

    o Improved scalability.
    
    o Added ability to process data sets larger than the memory of the 
    computer.
    
    o pcor() interface changed to allow the use of ff objects as input.
    
    o pcor() now returns a binary file as output instead of a text file. 

VERSION sprint prototype 0.0.3 
    
    o Introduction of the SPRINT library of parallel R functions.
    
    o Added ptest(), a function that check the library is installed and 
    operating properly.
    
    o Added pcor(), a parallel version of the Pearson correlation. pcor() 
    takes a 2D array as input and correlates each row with every other row.
    
==============================================================================
SPRINT Team

email: sprint@ed.ac.uk
http://www.r-sprint.org

Copyright  2012 The University of Edinburgh.
==============================================================================
   
