Re: [Teem-users] two pretty big changes in nrrd

[Gordon K. <gk...@cs...>]

hi,

I meant to paste in the listing of available nrrdKind values in my last 
message:

/*
******** nrrdKind enum
**
** The very cautious (and last?) step nrrd takes towards sementics.
**
** More of these may be added in the future, as when nrrd supports 
bricking.
**
** NB: The nrrdKindSize() function returns the suggested length for 
these.
*/
enum {
   nrrdKindUnknown,
   nrrdKindDomain,    /*  1: "It makes sense to resample along here" */
   nrrdKindList,      /*  2: "It makes NO sense to resample along here" 
*/
   nrrdKindStub,      /*  3: axis with one sample */
   nrrdKindScalar,    /*  4: same as a stub */
   nrrdKindComplex,   /*  5: real and imaginary components of a number */
   nrrdKind3Color,    /*  6: ANY 3-component color value */
   nrrdKind4Color,    /*  7: ANY 4-component color value */
   nrrdKind3Vector,   /*  8: any 3-vector */
   nrrdKind3Normal,   /*  9: 3-vectors which are assumed normalized */
   nrrdKind4Vector,   /* 10: any 4-vector */
   nrrdKind6Tensor,   /* 11: the 6 unique elements of a symmetric 3x3 
tensor */
   nrrdKind7Tensor,   /* 12: some scalar "confidence" or "valid" value, 
and
                             the 6 unique elements of a symmetric 3x3 
tensor */
   nrrdKind9Tensor,   /* 13: all 9 elements of a 3x3 tensor */
   nrrdKind9Matrix,   /* 14: all 9 elements of a 3x3 matrix */
   nrrdKindLast
};
#define NRRD_KIND_MAX   14

The suggested sizes (via nrrdKindSize()) for all of these is:
0: nrrdKindDomain, nrrdKindList
1: nrrdKindStub, nrrdKindScalar
2: nrrdKindComplex
the rest: pretty obvious

Gordon