Bug in vertical coordinates for line fit in rho-x plane

I believe there is a bug in function void getVertCoord at line 193 in reconstruct.cpp.

for (unsigned int i = 1; i < z_v.size(); i++) {
    if ((z_v[i] - z_v[i - 1]) * forward >= 0.) {
      dy_sq = tr.r * tr.r - (z_v[0] - tr.zc) * (z_v[0] - tr.zc);

dy_sq should be built using z_v[i] inside the loop. Currently, all y-coordinates end up being the same as the first one... It is strange that the fit converges nonetheless; maybe it's because this effect is smoothed when computing rho from z and y