functions in ylmdec.i -

             alm= accumulate(n, f, mesh, rcap, rtrack, pts, dylm, flux, nrpp)  
 
     Do N batches (approximately rays_per_batch each, default 13000) of  
     rays from file F, MESH = hydra_mesh(F), analyzing the symmetry of  
     a capsule of radius RCAP (all rays tangent to this sphere),  
     tracing the rays only down to radius RTRACK (>1.001*RCAP).  
     This routine is designed to be called several times successively  
     to get better statistics.  PTS, DYLM, FLUX, and NRPP are all updated.  
     PTS = unit vectors in directions of points on capsule surface  
           defaults to bucky(3,1): 1280 points with an intrinsic  
           fidelity of about 1% out to l=12  
           use bucky(4,1,domega), dylm= get_ylm(pts,lmax,list)*domega(-,..)  
           to get 5120 points and intrinsic fidelity of about 0.2% at l=12  
     DYLM = ylm*domega for PTS  
     FLUX = accumulated flux at each of PTS; units are power/area  
     NRPP = accumulated number of rays per point of capsule surface  
     ALM = 91 coefficients of Ylm up to l=12 corresponding to FLUX  
           l= [0, 1,1, 2,2,2, 3,3,3,3, 4,4,4,4,4, ...]  
           m= [0, 0,1, 0,1,2, 0,1,2,3, 0,1,2,3,4, ...]  

             get_ylm(pts, lmax, list)  
 
     returns values of Ylm at the 3-by-anything array of unit vectors  
     PTS, for all l and m up to LMAX.  The return value is  
     (lmax+1)*(lmax+2)/2-by-anything.  The LIST is also returned; its  
     value is [1, 2,2, 3,3,3, 4,4,4,4, ..., LMAX+1,LMAX+1,...,LMAX+1]  
     -- i copies of each integer i up to LMAX+1.  This can be useful  
     in conjunction with the histogram function; its values are l+1  
     for each element of the returned array's first index.  The m  
     values are [0, 0,1, 0,1,2, 0,1,2,3, 0,1,2,3,4, ...].  Negative m  
     values are omitted, since they would be the conjugates of m>0.  
       llist= list-1;  
       mlist= (!list(dif))(cum)-(list-2)*(list-1)/2;