----------------------------------------------------------------------------------
@MSGID:
<908fc019-0a18-4efb-b82e-1058793ba223n@googlegroups.com> e28fd26a
@REPLY: 1@dont-email.me> 3f6736ef
@REPLYADDR Farzad Tatar
<tatarfarzad1992@gmail.com>
@REPLYTO 2:5075/128 Farzad Tatar
@CHRS: CP866 2
@RFC: 1 0
@RFC-References:
<aa2d38fc-1df2-4703-b1b3-67adf45e2793n@googlegroups.com> <8a0d90c2-1912-41f9-8cf1-76c8f12698adn@googlegroups.com>
<c8ac7613-a6bc-4804-a668-40dbc9de519en@googlegroups.com> <26454e02-d247-4d86-a721-6166cdf11d56n@googlegroups.com>
<232d1483-dbdd-4833-8697-6e15686b90c7n@googlegroups.com> <klmb03F5oasU1@mid.individual.net>
<65c791c9-0095-42ed-aa69-9c2ba04b3caen@googlegroups.com> 1@dont-email.me>
<1fc4a61c-b795-4de4-a50f-9db95f208aa6n@googlegroups.com> 1@dont-email.me>
@RFC-Message-ID:
<908fc019-0a18-4efb-b82e-1058793ba223n@googlegroups.com>
@TZUTC: -0700
@PID: G2/1.0
@TID: FIDOGATE-5.12-ge4e8b94
> Rather than playing with the "far values", I think you should be
> looking at any close pairs to make sure that none of the underlying
> locations are not effectively identical. Otherwise you could try
> replacing the values on the diagonal of your matrix which should be
> being evaluated as one, by one plus a small number. The effect of this
> is to change the exact interpolation to an approximate interpolation.
> You may find this referred to as a "nugget effect".
>
> As to the singular value decomposition, this may just identify the fact
> that you have a difficult matrix. I don`t think it would help with
> matrix inversion, but there may be some other way to use the theory to
> do the interpolation.
>
> You say that you cannot change the size of RBF matrix, but this should
> be possible by splitting the interpolation problem into a set of
> overlapping sub-regions and doing interpolation for each. This is a
> generalised nearest-neighbour approach. Alternatively you could treat
> the idea as developing your understanding of what is going-on, by
> looking at the results for a 1 by 1 matrix, a 2 by 2 matrix,... etc.
Thank you for sharing your ideas. However, the problem is already
solved by changing the basis function. I changed the RBF basis function
from Gaussian to Multi-Quadratic and the problem got solved. I would have
tackled your methodology if I had not succeeded. Thank you anyway for your
help.
Kind regards,
Farzad
--- G2/1.0
* Origin: usenet.network (2:5075/128)
SEEN-BY: 5001/100 5005/49 5015/255 5019/40 5020/715
848 1042 4441 12000
SEEN-BY: 5030/49 1081 5058/104 5075/128
@PATH: 5075/128 5020/1042 4441