Question: spin coefficients using npspin

I need to calculate spin coefficients using a tetrad of null vectors, {l, n, m and m*} that I constructed for a certain metric. The brute tensor calculations are very lengthy, especially for this metric. On the other hand I have tried to use Maple (9.5) npspin; and the example given on the use of npspin. It's not clear from the example whether the entries (1,1),...(2,3)...etc are: 1) from setting the rows of the covariant vectors l, n, m and m* adjuscent to each other(?) to form some 4 X 4 matrix; or 2) from the metric g (elements) through dyadic combinations (1/2)ln+(1/2)nl-(1/2)mm*-(1/2)m*m that form it. 3) or some other combination. I tried #1, even though it didn't make sense, and the results for the spin coefficients were incorrect (based on my knowledge of the symmetries of the metric). #2 does not seem to be the case from the example. Can someone educate me on this, with a sample worksheet preferably for a well known metric (say Kerr)? Thanks.
Please Wait...