How can I program a function for a time series change-point test?

I'm attempting to write a function for a time series change-point test based on a Wilcoxon-Mann-Whitney type rank sum statistic (It's based on rank-order). You find the rank for each point in the time series (Xi) and compute the sum of the ranks at each point (i) from the beginning to that point (SRi) and from each of these an adjusted sum (SAi). SRi= symsum(R(j),j,i) (I hope that's the right syntax for summation) and SAi=abs((2*SRi)-(i*(n-1))). Then find the maximum of the SAi's (over points 1 to n-1). Suppose the maximum value of the SAi's occurs at point n1 and suppose W is the corresponding value of the sum of the ranks at n1 (i.e., SRn1); define the following: n2=n-n1; Wcrit=n1(n+1)/2 ; Sw=[n1*n2*(n+1)/12].^0.5.

I hope this is all making sense. The indata is time series data. Any thoughts? Essentially, I'm working with climate data and I want to identify change-points that may be indicative of an inhomogeneity.

Thanks