Subject: Re: Linear Prediction vs. First Point Scaling
From: "David S. Stephenson"
Date: Tue, 07 Jan 1997 18:07:37 +0100
Lee, Ji Hyun wrote:
>
> To my knowledge, if we don't half dwell time delay during data
> acqusition, first point is scaled with some factor, which is to
> eliminate ridges.
> After I predict some first points of time domain data with the linear
> precdiction method and replace the experimental data with them, which is
> right : I should scale the first point as above or I don't need to do
> such a thing after linear prediction?
>
> Happy New Year.
>
> Pharmacy, Seoul National University
I think you should still multiply the first point by 0.5, even afer
linear prediction, only then will the discrete transform approximate
the continuous transform (this assumes that the LAST point of your
fid is nearly zero). For a discussion see "The Fast Fourier Transform"
by E.O.Brigham.
All the best
David
--
David S. Stephenson
Institute of Organic Chemistry FON +49-89-5902-229
University of Munich FAX +49-89-5902-483
Karlstrasse 23 EMAIL dss@org.chemie.uni-muenchen.de
80333 Munich URL http://www.chemie.uni-muenchen.de/cicum
Subject: Re: The Name "Gauss" used to describe magnetic cannon.
From: "David S. Stephenson"
Date: Tue, 07 Jan 1997 17:59:56 +0100
Malcolm MacLeod wrote:
>
> "C. Liu II" wrote:
>
> >I'm not an expert on Magnetic Flux Density so perhaps someone can help me
> >out.
>
> >Question 1:
> >So... Is it proper to be calling this gun a "Gauss Gun"? (I did a history
> >search on several scientist named Gauss, so I do believe I know who it is
> >named after though not 100% sure)
> >A (computer) game that I'm testing uses a "Gauss" cannon as part of the
> >arsenal. However, the developer of the game decided that the Gauss cannon
> >was Gas compression based instead of Magnetic.
>
> >Question 2,
> >Between the two, which is correct? A "Gauss" cannon that uses gas or
> >magnetism?
> >Are either correct?
>
> "Gauss" relates specifically to magnetism, or magnetic fields.
>
> A gas-powered gun might be called a "Boyle" cannon...
A "Gauss" cannon is one in which the distribution of trajectories
is Gaussian - thus the employment of this nomenclature in a computer
game to emphasise that the thing doesn't always shoot straight!
--
David S. Stephenson
Institute of Organic Chemistry FON +49-89-5902-229
University of Munich FAX +49-89-5902-483
Karlstrasse 23 EMAIL dss@org.chemie.uni-muenchen.de
80333 Munich URL http://www.chemie.uni-muenchen.de/cicum
Subject: Functional MRI
From: mdp96akn@Sheffield.ac.uk (Amanda Nicholas)
Date: 8 Jan 1997 01:54:39 GMT
I am a first year postgrad in Clinical Neurology at Sheffield University,
UK, investigating the parts of the brain that malfunction is amnesia.
Having carried out a lot of reading in the last 3 months or so, I found
that a lot of research seemed inconclusive simply because it didn't
combine neuropsychological testing with functional and structural brain
images. I am determined not to fall into the same trap.
However, as yet, Sheffield University does not have access to a MRI
machine capable of functional imaging. We have connections with the Inst.
of Psych. in London, who have good facilities and I would be quite
prepared to travel down to London to carry out the scans.
The biggest problem with this is getting my subjects to London, or from
London to Sheffield. The neuropsychological testing facilities at the
University are very good, so ideally I would carry out this aspect of my
investigation here.
An alternative solution is to spend a year of my Phd in a place that has
both fMRI facilities and can provide supervision and facilities for the
neuropsych testing. This is probably the better option since subjects
would be local and thus available if any additional testing/scanning is
required.
I am also open to any other possibilities. My main priority at the moment
is to find facilities and funding for the functional scans. I am prepared
to consider any reasonable solutions to my problem, including teaching
undergrads, working in the labs in return for supervision etc etc.
Like a lot of people in my position, funding is also a major obstacle. I
am prepared to take out a loan to meet some of the costs if necessary, but
Mr bank manager will not give me enough to cover everything. If anyone
can help me out - either directly, or by giving me some names/addresses of
possible sources of funding, I'd be eternally grateful.
Return email address MDP96AKN@Sheffield.ac.uk
Thanks,
Amanda
--
Amanda Nicholas
The University of Sheffield, Department of Clinical Neurology, Floor 'N',
Royal Hallamshire Hospital, Glossop Road, Sheffield S10 2JF, UK.
Tel: 0114 2713158 Fax: 0114 2760095 EMail:mdp96akn@Sheffield.ac.uk
http://www.shef.ac.uk/uni/academic/A-C/cn/index.html
'`'`'`'`'`'`'`'`'`'`'`'`'`''`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`
sci.psychology.research is a moderated newsgroup.
Before submitting an article, please read the guidelines which are posted
here bimonthly or the charter on the web at http://www.grohol.com/spr/
Submissions are acknowledged automatically.
Subject: Re: Spectra conversion
From: jajones@ermine.ox.ac.uk (Jonathan Jones)
Date: 8 Jan 1997 10:42:10 GMT
In article <32D28E9E.7097@kemi.slu.se>,
Nils Nyberg wrote:
>Andre Pampel wrote:
>> can anybody help me with a program that converts 2D spectra (Bruker and
>> Felix format) in
>> Matlab-matrices and/or vice versa?
>
>We've done this conversion manually a few times. From Bruker format via
>the jcamp-dx ascii format to matlab. It works but it takes some time. So,
>I'm also interested in a program which accomplish this.
I quite frequently convert felix matrices to raw binary data (ie, just
the values stored as 4 byte binary floats, reading along each row from
bottom left to top right). It can't be too hard to get from that to
Matlab format.
Felix matrices are stored as 16K bytes of header, followed by the data
stored as 4 byte binary floats in a submatrix format. So all you have to
do is extract the details of the submatrix format from the header,
read in the data (for a program in C use the fread() function),
rearrange it, and then write it out (for a program in C use the
fwrite() function to get raw binary data). Even better you can skip
the rearrangement step by reading the data into the right bits of the
data matrix directly.
The submatrix format is obtained as follows. Suppose the matrix has
n1 by n2 points; these are stored as a series of nb1 by nb2 submatrix
blocks, each of size ns1 by ns2 (clearly n1=nb1*ns1, n2=nb2*ns2).
The blocks are stored starting with the block in the bottom left,
then the one next on the right, and so on, ending with the top right
block. The data within a block is stored running from bottom left to
top right as you might exepect.
To get the values of n1 etc, read the header as 4 byte binary integers
into an integer array (using fread()), and then
n1=ipar[22]; n2=ipar[23];
nb1=ipar[24]; nb2=ipar[25];
ns1=ipar[28]; ns2=ipar[29];
The "rearrangement on the fly" step can be done by reading in as follows:
for (i=0;i
Return to Top
Downloaded by WWW Programs
Byron Palmer
