*******************************************************************


At 03:20 PM 3/9/2005 -0900, you wrote:
Is a copy of TeX available to students through the university?
As I'm sure you are aware, wordpad doesn't turn out to be the best
application for typing mathematics.

DESPERATELY SEEKING GOOD TYPESETTING



To: DESPERATELY
Subject: TeX

It's free.  For Windows you will find two links on my
home page, one to
http://miktex.org/
for a TeX distribution and one to
http://www.winedt.com/
for a really good editor (shareware).  Both of these are Windows
things; for Linux there are even better free options but I am not
the one to ask.

Ed



********************************************************************


At 12:26 PM 3/10/2005 -0900, you wrote:
I had a Matlab question unrelated to the class that I was hoping
you could help me with. I have a dataset that contains glacier
elevation h at a distance x along a flow line. In order to use
this data to extract thickness estimates I need to have data points
that are spaced a specified distance and I was using a cubic spline
interpolation to get these points, but this is giving me problems,
and I would like to take the dataset and just get a linear
interpolation between each data point and then sample this
interpolation at the points I need for proper spacing. Is there a
built-in Matlab function which will perform this or do I need to
take the time to write my own code?


Thanks

GLACIOLOGIST STUCK IN ICE



To: STUCK
Subject: Re: matlab help

Yes.  It's  interp1, as in

>> x=[0.0 0.1 0.4 0.7 1.0 1.5]
>> H=[45 34 44 50 50 63]
>> xx=0:0.01:1.5;
>> HH=interp1(x,H,xx);
>> plot(x,H,'o',xx,HH,'.')

This gives values of  H  on the regularly spaced grid with  dx =0.01.
The last command shows that linear interpolation has been used.
The same command   interp1  can be used for cubic spline interpolation:

>> HH=interp1(x,H,xx,'spline');
>> plot(x,H,'o',xx,HH,'.')

Other options are possible.  See also
spline
interp2
interp3
for related functions.


Ed


PS:  To first approximation, NEVER write your own code.
In fact, the following equation applies:

F(desired code) = F(MATLAB builtin)
  + F'(MATLAB builtin) (doable at command line)
  + (1/2) F''(MATLAB builtin) (doable at command line)^2
  + (1/3!) F'''(your own code using as much MATLAB builtin as possible)
        (doable at command line)^3