The following code is a function (performance-critical) to compute tied ranks of a vector:

mergeSort(x,inds,ci);
//a sort function to sort vector x of length ci, also returns keys (inds) of x.

int tj=0;
double xi=x[0];

for (int j = 1; j < ci; ++j)
{
    if (x[j] > xi)
    {
        double rankvalue = 0.5 * (j - 1 + tj);
        for (int k = tj; k < j; ++k)
        {
            ranks[inds[k]]=rankvalue;
        };

        tj = j;
        xi = x[j];
    };      
};

double rankvalue = 0.5 * (ci - 1 + tj);

for (int k = tj; k < ci; ++k)
{
    ranks[inds[k]]=rankvalue;
};

The problem is, the supposed performance bottleneck mergeSort(), which is O(NlogN) is several times faster than the other part of codes (which is O(N)), which suggests there is room for huge improvment with the other part of the codes, any advices?