home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Aminet 5
/
Aminet 5 - March 1995.iso
/
Aminet
/
misc
/
sci
/
MLSv1_7.lha
/
MLS_V1.5.readme
< prev
next >
Wrap
Text File
|
1993-08-01
|
3KB
|
84 lines
MLS v1.5 (Method of Least Squares)
Written By Brian Olson
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
This is ShareWare
if you find this program useful
please send at least $10 to the Author
to encourage Updates.
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Authors Address:
Brian Olson
11 Limetree Lane
Liverpool, NY 13090
USA
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
The Method of Least Squares is a graphical analysis program that
determines the relationship between two sets of data. Plotting the two
sets of data against each other and drawing a line which touches, or comes
close to, all of the data points (The Best Fit Line) then the equation of
that line would tell you how the two sets of data are related.
Once the relationship is determined you could extrapolate to find
ALL possible data points.
However drawing the Best Fit line can be very difficult, expecially
if the data doesn't represent a straight line. The Method of Least Squares
makes this process easy, it will determine the Best Fit Line for you.
There are, however, many repetitive calculations in calculating the
Best Fit Line, therefore I decided to write a program to do it for me.
To make life even easier for the User, I installed a couple of
preset relationships into the program. They are
Relationship Representative
Equation
( a and b are constants
to be determines )
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Linear (Definatly Need this one) Y = aX + b
Power Y = aX^b
Exponential Y = aEXP(b)
Logarithmic X-Axis Y = aLOG10(bX)
Logarithmic Y-Axis LOG10(Y) = aX + b
Logarithmic X & Y Axis LOG10(Y) = aLOG10(X) + b
The first thing you need to do is make a "points" file. Just load up your
favorite text editor and enter in the X and Y data points, one set of X and
Y points per line
{ex.}
1.32 2.42
2.43 3.53
3.46 4.57
it's as easy as that. Next load up the MLS program and click in the points
filename gadget, enter the name of the file you just created, hit Enter and
then click on the LOAD Button. The Status window will tell you what's going
on.
Once the points are loaded, they are plotted in the bix box on the
screen. The X and Y Axis Ranges are shown just below the box. By looking at
the positions of the points it's possible to determine the type of
relationship. Using this guess, click on the little button next to the
relationship you believe these points to represent. Now click on CALCULATE.
The resulting line will be drawn. If it doesn't come close to looking like
the Best Fit line you probably selected the wrong Regression Mode. If it
DOES look like the best fit line, you can note the equation in the status
line, or if you want a more elaborate display of data click on the SAVE
button. A data file with the same name are your points file but with a
.REG extenstion will be created.
If the Regression Mode of the points is not one of the Preset mode
then you have to decide what it represents, and then convert it to a Linear
equation, converting the points along with it. And run the program and
LINEAR with your NEW points. I won't really go into this process.
Well. I guess that's about it... Enjoy...