You have chosen to sponsor your bid up to a maximum amount of .
I have written a program in JustBasic (a free dialect of Basic) that inputs a comma delimited matrix of numbers, and uses a computationally-intensive procedure to compute a coefficient alpha, a statistic measuring the internal consistency of the numbers. I am looking for a programming freelancer who is very familiar with the statistical computing package R, who can rewrite my program in such a way as to be readily accessible to those who use R. The website for R states, "Much of the system is itself written in the R dialect of S, which makes it easy for users to follow the algorithmic choices made. For computationally-intensive tasks, C, C++ and Fortran code can be linked and called at run time." You can help me to decide whether the code should be written in the "R dialect of S," or in C++ and linked, or both. I want someone familiar enough with R to be able to tell me exactly how to do this with R, not just to produce a C++ program. The original program is very highly commented, and I believe it should be easy to follow, even if you don't program in JustBasic. The procedure in the original code consists in taking the number of "items" in the "scale" (i.e. the number of variables that form the rows of the matrix) and randomly sorting them into two halves. Then the program computes the mean of each of these two subsets, for each case, treating missing values by basing the mean only on those values that are nonmissing. Then when done with all the cases, the program computes a Pearson correlation between the means of the two subsets. Then the program randomly divides the items into two subsets again, and does the same thing all over again, and continues for many random subsets. The program "steps up" the correlation coeeficients using the Spearman-Brown formula. The program you write will be in the public domain or the GNU equivalent of the public domain, so your work will be "for hire" and you won't retain any copyright to it. I think that someone who is very familiar with the programs involved should be able to do this task very quickly. The original code is not very long.