LNSD is the jCommon module of jLab.

  LNSD  Last non-singleton dimension of an array.
   
    LNSD(X) returns the number of the last non-singleton dimension of X.
 
    This provides a useful definition of the dimensionality of X.  Unlike 
    Matlab's NDIMS, which thinks that a column vector and a scalar both 
    have dimension 2, LNSD defines the dimension of a scalar to be zero and
    that of a column vector to be one, while a row vector has an LNSD of 2.
   
    Note that LNSD(X) changes if X is permuted. 
   
    [N1,N2,...NM]=LNSD(X1,X2,...XM) returns the dimensions of multiple 
    input arguments.  If zero or one output arguments are given, a single 
    row array [N1 N2 ...NM] is output.
    _________________________________________________________________
    This is part of JLAB --- type 'help jlab' for more information
    (C) 2002--2015 J.M. Lilly --- type 'help jlab_license' for details