This is a sample summary file from the first part of Lab 1.  This
work was done on the version of Matlab on unixs.cis.pitt.edu,
so some of the responses might be slightly different than on the
version you are using.  This summary is written in the first person, 
but you can use the third person if that makes you more comfortable.


Exercise 1:
The purpose of this exercise is illustrate editing a downloaded file.

1.1  I started up Matlab and used the "Current Directory" window to
create the directory "math2071" and then the directory "lab01"
inside "math2071."  Then I typed in the commands

diary diary.txt
% Lab 1 summary.  M. M. Sussman

the last is just a comment that will show up in the diary.txt file.

1.2  From the browser, I right-clicked on the file demoscript.m, 
and followed directions to download a file to "math2071/lab01".
This file appeared in the "Current Directory" file listing and
also when I did a dir command and got
.            ..           demoscript.m diary.txt    exercise1

1.3  I was able to edit the file with the "edit demoscript.m" command.

1.4  I executed the file from the Tools-->Run menu choice.  The
result was (after eliminating blank lines):
x =
    1.3000
xsquared =
    1.6900
p =
    1.8540
ans =
    3.1416
adjacent =
    0.7071
z =
   3.0000 + 2.0000i
zsquared =
   5.0000 +12.0000i
logz =
   1.2825 + 0.5880i
ans =
The following is a list of cube roots.
cuberoots =

  Columns 1 through 7

    1.0000    1.2599    1.4422    1.5874    1.7100    1.8171    1.9129

  Columns 8 through 10

    2.0000    2.0801    2.1544

Exercise 2.  The point of this exercise is to demonstrate the
debugging trace utility.

2.1  I found the menu picks: Debug-->"Stop if Error" and
Debug-->"Stop if NaN or Inf" and activated them.

2.2  I found the statement bad=1/(x-1); near the bottom of the
file, took away the percent character, and saved the file.

2.3,2.4  I re-ran the script.  I saw the same output as previously,
but it ended with 
Warning: Divide by zero.
> In /afs/pitt.edu/home/m/m/mms125/lab01/demoscript.m at line 54
bad =
   Inf
and a blue arrow appeared next to the bad= statement in the
edit window.  The special debugging prompt "K>>" appeared in
the command window.

2.5  I hovered the mouse pointer over "x" and saw its value = 1.

2.7  I exited debugging mode with the Debug-->"Quit Debugging" menu
choice.

Exercise 3.  The point of this exercise is to illustrate how to
trace execution using the debugger.

3.1  I typed "clear".

3.2  I set a breakpoint on the first executable line.  A red dot
appeared there.

3.3  I selected "Run" from the debug menu.  An arrow appeared, pointing
to the line with the red dot.

3.4,5  I used the "Step" button (First of five buttons in the last
group of buttons at the top of the "Edit" window.  It has a curved
arrow on it and an image of a single page.) to step through the
file.  As new variables were defined, they appeared in the
"Current Workspace" windowpane, and results appeared in the 
"Command" windowpane.

4.6  The loop to compute cube roots was tedious, so I put another
breakpoint just after the loop and used the "Continue" button
(I could have used a menu pick) to jump directly to the new
breakpoint.

4.7  I used the "Exit Debug Mode" button.

THE FOLLOWING IS A SUMMARY OF AN EXERCISE SIMILAR TO ONE USED FOR
THE FIRST SEMESTER COURSE, MATH 2070.  IT IS INCLUDED HERE AS AN
EXAMPLE OF A MORE TECHNICAL EXERCISE.

5. The point of this exercise is to illustrate using the Matlab
"polyfit" command to construction a polynomial curve approximately 
through given data points.

5.1 I downloaded the data points from the web page and read them
into Matlab.  The result was two vectors, xdata and ydata, so
that each data point is represented as (xdata(i),ydata(i)) for
i between 1 and 10.

5.2 I used the command "c=polyfit(xdata,ydata,3)" to find the coefficients
of a parabola that passes near the given data points.  These coefficients
are  given by  c=[1.1 3.2 -5 12].

5.3 I constructed a set of abscissas for plotting.  Since the data points
lie between x=0 and x=1, I chose 100 points between x=-.5 and x=+1.5
with the command "xplot=linspace(-.5,1.5,100);".  100 points will make a
nice, smooth curve.  I used "yplot=polyval(c,xplot);" to find ordinates
corresponding to the abscissas xplot.

5.4 I plotted both the given data points and the parabola together
using the Matlab command "plot(xdata,ydata,'*',xplot,yplot)".  I used
the File-->Export command to save the plot as a "Jpeg image"
named comparison.jpg, included along with this summary.  The line
passes very nearly through the "*" symbols representing the given
data points.

$Date: 2007/12/20 16:07:44 $
M. M. Sussman