POLYSMOOTH_KERNEL Returns the weighting kernel employed by POLYSMOOTH. POLYSMOOTH_KERNEL is a low-level function called by POLYSMOOTH. W=POLYSMOOTH_KERNEL(DS,[],B,KERNSTR) where DS is an array of distances and B is a bandwidth, returns the value of the weighting function specified by the string KERNSTR at distances DS. See POLYSMOOTH for valid choices of the string KERNSTR. B is either a scalar, or a matrix the same size as DS(:,:,1). W=POLYSMOOTH_KERNEL(DS,[],B,L) where L is a scalar uses a Gaussian with a standard deviation set to B/L, as in W=EXP(-1/2*(L*DS/B)^2). W=POLYSMOOTH_KERNEL(DS,[],B,K) where K is an array takes K to be a custom kernel, with K(1) corresponding to DS/B=0 and K(end) to DS/B=1, interpolates within K to find the weight W at any distance. W=POLYSMOOTH_KERNEL(T,[],TAU,KERNSTR) works in just the same way if T is a time deviation and TAU is a temporal bandwidth. W=POLYSMOOTH_KERNEL(DS,WS,B,KERNSTR) optionally multiplies W, after computed as described above, by the additional weights given in WS. See also POLYSMOOTH. Usage: W=polysmooth_kernel(d,[],B,kernstr); W=polysmooth_kernel(d,ws,B,kernstr); __________________________________________________________________ This is part of JLAB --- type 'help jlab' for more information (C) 2018--2020 J.M. Lilly --- type 'help jlab_license' for details