Category : Science and Education
Archive   : DA16.ZIP
Filename : DA.DOC

****************************************
* *
* The DATA ANALYST *
* version 1.6 *
****************************************
(c) Copyright 1989, All rights reserved.

by

Kin-Peng Sin
P.O. Box 9316
University MS 38677



License agreement:

This manual and the accompanying program may be copied and
distributed only in the following manner:

i) given to friends
ii) uploaded to bulletin boards
iii) included in the library of a nonprofit
computer club.

The sale of this program by any establishment either separately or
as part of a 'library' diskette is strictly forbidden.



Shareware Notice:

This is not a public domain program. If you find this program to
be of use to you, please register by sending a contribution.
Contributors who send $20 or more will receive a free upgrade of
the next release. Further upgrades for registered users will cost
$5 to cover the cost of media, postage, and handling. Your
contributions will encourage me to expand the capabilities of the
program to include extended EGA and VGA graphics support,
additional statistical functions, improved printer and plotter
support, etc.



DISCLAIMER OF WARRANTY

THIS PROGRAM AND MANUAL ARE PROVIDED "AS IS" WITHOUT WARRANTY OF
ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED
TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS OF
PURPOSE. THE USER OF THIS SOFTWARE AND MANUAL ASSUMES ALL RISKS.



The Data Analyst. ver 1.6



TABLE OF CONTENTS

Topic page


1. Introduction . . . . . . . . . . . . . . . . . . . . . . . 1
Description
Machine and software requirements
Acronyms used in the program

2. Using the Data Analyst . . . . . . . . . . . . . . . . . . 3

3. Data File Formats . . . . . . . . . . . . . . . . . . . . 5

4. The Graphics Mode . . . . . . . . . . . . . . . . . . . . 6
Summary of the graphics mode functions
Cursor functions

5. Using the Built-In Editor . . . . . . . . . . . . . . . . 9

6. Changing the Graph Display Options . . . . . . . . . . . . 10

7. Notes on Statistical Functions . . . . . . . . . . . . . . 11

8. Errors and Error Messages . . . . . . . . . . . . . . . . 12
More notes on labeling

9. Notes on Printing and Plotting . . . . . . . . . . . . . . 13

10. Tutorial . . . . . . . . . . . . . . . . . . . . . . . . . 14

11. User Feedback . . . . . . . . . . . . . . . . . . . . . . 17

12. General Facts . . . . . . . . . . . . . . . . . . . . . . 17


- Registration/Order Form -



The Data Analyst. ver 1.6 page 1



1. Introduction

Description

The Data Analyst is a program designed to let users analyze data
in x-y pairs by simple statistical tests and regression
techniques. Pull-down menus, a built-in editor, and a flexible
graphical interface allow the user to manipulate and view data in
a variety of ways. Features include the capability to toggle
between Log and Non-log scales 'on-the-fly' and performing
regression analysis selectively on portions of a data set. A few
popular printers and HP-GL compatible plotters are supported to
reproduce on-screen graphs.

Summary of the program's features:

- Generate regression curves
Polynomial fit (up to 9th order)
Least Squares, Log, Power, Exponential, and
Binomial Power fit.
- Perform regression analysis on all or a portion of a
data set.
- Obtain basic statistical information on a set of data.
- Determine slope of two user defined points on-screen.
- Scale data (add or multiply x,y points by a
constant/factor)
- Coordinate transformation. (Log (any base), Semi-Log,
and Cartesian system)
- Transposing x-y points.
- User defined titles and labels. Labels can be placed
anywhere in the plot.
- Built-in data editor, HP-GL compatible plotter drivers.
- Generate presentation quality plots of regression lines
with or without data points.
- Convenient graph editor.

This program is targeted towards scientists and engineers who require
the means to perform basic statistics and regression analyses on
their data. The design philosophy during program development was to
create a compact package that is both easy to learn and use. Most
functions offered in the program can be accessed with just a few
keystrokes. The simplicity of the user interface in the Data Analyst
permits users to master the program quickly. This program will not
persuade power users to part with their complex statistical packages.
However, the worker in a laboratory or even a business environment
might find this program useful in analyzing and graphing relatively
small data sets quickly and efficiently.



The Data Analyst. ver 1.6 page 2



Machine and software requirements

- IBM-PC, XT, AT, or 100% compatibles
(preferably with a numeric co-processor)
- Color Graphics Adaptor
- 512k RAM or more
- One disk drive
- DOS 2.0 or higher
- Text editor (optional but recommended)


Acronyms used in the documentation and program:

ADS - Active Data Set: the data set that any selected function in
the program will act on.
NDS - total Number of Data Sets: the total number of x-y data
sets in memory. (maximum = 5)
Lg-B - Log Base


Other facts:
Data set limit - The Data Analyst currently allows up to 500 x-y
pairs per data set.

Note: The key and the Enter key are one and the same.



The Data Analyst. ver 1.6 page 3



2. Using the Data Analyst

It is imperative that the user reads all of the information given
in this manual to understand fully the capabilities of the
program. Although the manual may seem to be long, the program can
be easily mastered within a short time. Some users may find it
quicker to skip ahead to the Tutorial section and start working
with the program and then returning to this section.



The program is started from the DOS prompt:

(d):>da

The program title will appear after a few seconds. Depressing the
appropriate keys will bring the user to options on the main menu.
A pull-down menu will appear for each main option. Use the left or
right arrow keys to select different options along the bar. The
up or down arrow keys allow the selection of options within each
pull-down menu. A description of each of the functions in the
pull-down menus follow.

Bar Option - FILE
Pull-Down options
Load - load a data file, (if a file is already in
memory, you will be prompted to cancel your
pick or replace the file with a new one.)
The graph will automatically be scaled to
fit the FIRST data set initially.
Edit - edit a datafile. (after entering the last
data point, hit the key to ensure the
point has been accepted by the editor.
Save - saves a data file to disk.
Help - displays a help screen.
Shell - exit temporarily to the
operating system.
New - erases current data sets from memory and
resets all variables to the default settings.
Quit - quit the program.

Bar Option - STATISTICS
Pull-Down options
Regression - select the desired type of regression
analysis to be performed on data
while in the graphics mode.
Basic Stats - runs simple statistical tests on
selected data set.



The Data Analyst. ver 1.6 page 4



Bar Option - GRAPH
Pull-Down options
View - enter graphics mode.
Options - modifies scaling (Cartesian and Log),
labels, additive (cx,cy), and
multiplicative (mx,my) factors.
Defaults - displays the current settings to be
used in the graphics mode.


Bar Option - PRINT
Pull-Down options
DMPrinter - print graph on dot-matrix printer.
Plotter - print graph on pen plotter.
Options - configures the type of output on the
dot-matrix printer.

Bar Option - CONFIGURE
Pull-Down options
Log Drive - lets user set drive containing data
files.
DMPrinter - select appropriate dot matrix printer
which matches user's hardware settings.
Plotter - select pen plotter corresponding to
user's hardware settings.
Com Port - select communication port to pen
plotter.
Settings - shows current system settings.



The Data Analyst. ver 1.6 page 5


3. Data File Formats

Three file formats are used. The first line of data in all three
formats must contain the following information:

[file code], [number of data sets]

A file code of 1 is used when the number of x-y pairs in each data
set is the same. Set the file code to 2 when the numbers are not
equal.

example 1: [Format 1] 3 data sets, each containing 5 x-y pairs

1,3 <- file code = 1, 3 data sets
x11,y11,x21,y21,x31,y31
x12,y12,x22,y22,x32,y32
x13,y13,x23,y23,x33,y33
x14,y14,x24,y24,x34,y34
x15,y15,x25,y25,x35,y35

example 2: [Format 2] 2 data sets, data set #1 has 2 x-y pairs
and data set #2 has 3 x-y pairs.

2,2 <- file code = 2, 2 data sets
2,3 <- line indicating number of points in each set.
x11,y11 In this case, data set #1 has 2 points and
x12,y12 data set #2 has 3 points.
x21,y21
x22,y22
x23,y23

example 3: [Format 3] 3 data sets, each containing 5 x-y pairs.
The x value of each x-y pair is the same as the x
values of the corresponding x-y pairs in other data
sets.

3,3 <- file code = 3, 3 data sets
x11,y11,y21,y31
x12,y12,y22,y32
x13,y13,y23,y33
x14,y14,y24,y34
x15,y15,y25,y35

Terminators may either be commas or spaces.

Notation :
x1- and y1- are data points in the first data set and
x2- and y2- are data points in the second data set, etc.

Note : If the built-in editor is used for data entry, the above
guidelines are followed automatically when the file is
saved. Files of format type 3 read by the Data Analyst
are saved as format type 2. Files created in spreadsheets
or other programs with the above formats and saved as text
(ASCII) files will also work.


The Data Analyst. ver 1.6 page 6



4. The Graphics Mode

Selecting the VIEW option in the GRAPH menu will invoke the
graphics mode. In this mode, a graph will be displayed according
to the specifications shown in the DEFAULTS selection of the GRAPH
menu. On the right side of the graph are five status boxes
indicating the various options that are in effect. The boxes
labeled X-Log and Y-Log show if either of the X or Y axis is in
the Log mode. A box labeled Lg-B displays the current Log base.
Below it are two boxes (NDS and ADS) indicating the number of data
sets in memory and the current active data set. The graph title
is not displayed in this mode (except in the preview option).

Situated below the graph are two text lines. The upper box
comprise (from left to right) the X-Y coordinate of the cursor
location, a message box, and the current user-selected regression
mode.

Functions that can be performed in the graphics mode:

* Display the help screen. This is accomplished by
depressing the F1 key.

* Expand the graph to cover the entire screen. [F2 key]
Use this to preview the finished graph.

* Insert labels anywhere within the graph. To do this, use
the arrow keys to direct the small cursor located near the
bottom of the graph to the desired location. Depress the
F3 key to enable text input. Hitting the key will
terminate the option. The program allows a maximum of 15
on-graph labels. The string starts at the point of the
cursor when F3 was pressed in the final output. Labels are
treated as special data points by the program with its
starting location defined at the point of the cursor when
F3 is pressed. The label itself is treated as a data point
symbol. If the user makes an error while typing the label,
simply hit and [C] to clear the label.

* Toggle between Log and Non-log scales. Use the F4, X, or Y
key to do this. Should any data set contain negative values,
the function is disabled.

* Display or Hide the grid. [F5 key]

* Display results of regression analysis. [F6] A text screen
with the appropriate data will be displayed. Note that the
user must have performed the analysis first before this
screen can be displayed.

* Determine the slope or gradient between two user-specified
points. The two points are set by pressing the F7 key when
the cursor is at the desired locations.


The Data Analyst. ver 1.6 page 7



* Perform regression analysis selectively on portions of a
data set. A segment of the data set must be determined
before this can be achieved (a segment is defined as two
points along the X-axis). This is done by first moving the
cursor to one location and hitting the F8 key. Specify the
end of the segment by repeating the above procedure. A
regression line should appear after a few seconds. Hit the
F6 key to view the results. In general, a minimum of TWO
data points must be contained within the segment to enable
the analysis. REMEMBER: The analysis is performed on the
Active Data Set (ADS). Nothing will appear to happen if
the ADS is not within the limits of the current graph
boundaries. 2nd REMINDER: Rescale the graph to fit the ADS
with the [S] key. Select the ADS using the [I] key.

* Obtain a raw screen dump of the graph. [F9]
GRAPHICS.COM must be installed first.

* Alter the speed and accuracy of the cursor. [F10]

* Change the Active Data Set. Press the [I] key until the
desired data set number is shown on the right side of the
screen.

* Scale the graph to fit the Active Data Set. [S]
* Scale graph to fit all data sets. [A]

* Perform regression analysis on the entire Active Data Set.
[F]. To view the regression line make sure the ADS is
within the current graph limits.

* Delete the most recently entered user-defined label. [C]

* Redraw the graph. [R] Use this when the graph becomes
cluttered with regression lines, labels, or gradient lines.

* Toggle between decimal and exponential tic labels. [E]
This function becomes useful in the event that the length
of the tic mark labels become excessively long. This
function is automatically used if the magnitude of the
numbers become large. Always preview with the F2 key to
ensure that all labels are properly displayed. If they
overlap or extend beyond screen limits, reduce the number
of tic marks and use exponential labels.

* Redraw the most recently computed regression line for each
data set. [P]

* Erase all of the computed regression lines from memory. [U]

* Display the legend. [L] This will not show up in the
preview mode or in the finished graph.


The Data Analyst. ver 1.6 page 8



Summary of the graphics mode functions:

F1 - displays a help screen
F2 - expands plot to full-screen
F3 - inserts on-graph label at current cursor position
F4 - toggles between Log and Non-Log scales
F5 - hide or show grid
F6 - displays regression results of most recent analysis
F7 - determine the gradient between two points on the graph
(use the cursor to pick desired points)
F8 - pick portion of data set to perform regression analysis
(use the cursor to pick desired points)
F9 - screen dump of plot (GRAPHICS.COM must be installed)
F10 - slows down or speeds up cursor
Spacebar - cycles thru regression options (ver 1.2 upwards)

- return to the main menu
[I] - change the active data set
[C] - clears most recently entered label
[R] - refreshes the graphics screen
[S] - scales graph to fit selected data set
[A] - scales graph to fit all data sets
[F] - performs regression analysis on ALL points of
the selected data set
[X] - toggles between Log and Non-Log scales on the X axis
[Y] - toggles between Log and Non-Log scales on the Y axis
[L] - display the legend
[P] - Redisplay regression lines
[U] - Remove regression lines
[E] - toggles between decimal and exponential labels
[T] - transposes the x and y values of the active data set
(make sure that any regression lines drawn prior
to the transpose operation are removed using the U
key)

Arrow keys - move cursor


Functions of the Cursor.

The cursor is a multipurpose tool used to do the following:

- determine the starting location of a label.
- specify the starting and ending points of a segment of
a data set for selective regression analysis.
- specify two points on the graph for determining the
slope.
- determine more precisely the X-Y coordinate of a data
point on the graph.



The Data Analyst. ver 1.6 page 9



5. Using the Built-In Editor

Selecting EDIT from the FILE menu will bring up the editor. If
no data is present in memory, the user will be prompted to edit
the 1st data set. If data is present, then a second menu appears
letting the user select the data set to be edited. When there is
one data set already in memory, the user can edit either data set
#1 or #2. If two data sets are in memory, data sets #1, #2, and
#3 can be edited. In short, when data is present, the user can
add additional data sets by editing the data set designated one
number higher than the total number of data sets in memory.

The editor can only be used to edit numbers within the range
-99999 to 999999. It will not be activated when data points
exceeding these limits are detected. User's wishing to
analyze data outside these limits will have to use a more
versatile text editor to enter data.

The editor functions as follows:-

* The editor starts up in insert mode. Moving the cursor
using the left or right arrow keys will cause it to
switch to the overwrite mode.

* Alt-X will delete the current hi-lighted data point.

* Alt-C will zero the current entry.

* Because the data set is automatically sorted by X-values
upon exit, no Insert feature is necessary.

* Due to the way in which the editor detects the number of
data points present, if the last data point entered is
[0,0], it will not be recognized as valid data.
However, a [0,0] point entered in the midst of other
nonzero points is recognized.

Note: A data set must consist of at least two points.



The Data Analyst. ver 1.6 page 10



6. Changing the Graph Display Options

Selecting OPTIONS from the GRAPH menu will enable the user to
change various display options. The options that can be altered
are:

* Main title, X and Y axis labels

* Number of tic marks on the X and Y axis in the cartesian
mode. (Maximum of 10 tics, minimum of 1 tic)
(default = 6)

* Display or hide the data point symbols (OFF/ON)
(default = ON)

* Connect the data points with straight lines (OFF/ON)
(default = OFF)

* Manually scale the X and Y axis. The maximum and
minimum allowable user-specified values at present are
999999 and -99999 respectively. Users are limited to
using the autoscale feature when data values exceed the
above limits.

* Change the Log base. (default = 10)

* Change the starting Log cycle. (default = -1)

* Change the number of Log cycles in the X and Y axis.
The maximum number of cycles allowed is five.
(default = 4)

* Multiply each x and y point in a data set by a factor.
(MX, MY) (default = 1, limits : -999 to 9999)

* Add each x and y point in a data set by a constant. (CX,
CY) (default = 0, limits : -999 to 9999)


Note: For some of the options, the user is forced to enter valid
data before being allowed to leave the field. (e.g. if the
minimum x limit is set to be larger than the maximum x-axis
limit, the program will sound a tone and prompt for a valid
entry)



The Data Analyst. ver 1.6 page 11



7. Notes on Statistical Functions

- The acronyms used in the window containing basic
statistical information on a data set are as follows:

STD X - standard deviation of the x values in the set
STD Y - standard deviation of the y values in the set
min X - minimum x value in the set
max X - maximum x value in the set
min Y - minimum y value in the set
max Y - maximum y value in the set

- The results displayed in the Basic Statistics
window has taken in account any multiplicative or
cumulative factors modified in the OPTIONS
selection of the GRAPH menu.

- When using the Binomial Power fit, users have the option
to enter the desired power values.

- When a segment of a data defined by the user contains
less than the required number of data points for a
polynomial fit, the message [Insufficient points] will
be shown in the message box.

- In certain cases, some data sets contain data that will
result in the creation of an ill-conditioned matrix when
users attempt to fit a polynomial curve to the data. In
such cases, the message [Singular matrix!] will be
displayed briefly in the message box. The resulting
regression line (if any) and regression results will be
suspect and should be ignored.

- A data set (or a data segment) must contain a minimum of
two points for any type of regression analysis to take
place.



The Data Analyst. ver 1.6 page 12



8. Errors and Error Messages

Error messages might appear when an error occurs in any of the
three modes:

[1] Loading or saving files. The message [File I/O error]
will appear under these conditions:

i. when the user attempts to load a file which is not a
recognized data file.
ii. disk error. write protected diskette, open disk
drive door, etc.
iii. when the user attempts to save a file to a
nonexistent directory.

[2] Running BASIC STATS on a bad data set. The message
[Warning! Illegal data] will appear if a data set
contains insufficient data points or data which is
inappropriate for analysis.

[3] When the user is in the graphics mode, the most common
error occurs (after rescaling the graph to fit a
different data set) when the user-defined labels are too
lengthy to fit within the limits of the graph. The
offending labels will be partially displayed or be seen
to overlap areas outside the graph. The solution is to
scale the graph manually to accommodate the long label.
Labels are treated as long data point symbols. An
alternative solution to this problem is not to rescale
the graph after labels have been entered.


More notes on labeling:

The label mode cannot be activated if the cursor is too
close to the right and bottom boundary of the graph.
This mode is automatically deactivated before any text
overwrites the right edge of the graph. Be wary of
labels which have been defined to start at a negative x
or y coordinate value. Switching to the log mode will
cause these labels to be erratically placed.



The Data Analyst. ver 1.6 page 13



9. Notes on Printing and Plotting

- The HP7475A and the HP ColorPro should be set at
9600,S,7,1

- For the HP ColorPro to work, it must contain the
graphics module.

- The HP Draftmaster-I should be set at 9600,N,8,1
The same applies to the Draftpro EXL.

- Wrong communications settings will result in garbled
plotter output.

- Owners of architectural-size plotters which emulate the
HP-Draftmaster can use the HP-Draftmaster option in the
plotter selection menu.

- Use 8.5"x11" paper on all plotters.

- All dot-matrix printer drivers were written for 80-
column printers.

- Graphics output to dot-matrix printers is not
presentation quality. This feature will be improved
upon in future releases of this software if there is
sufficient user demand.

- Owners of Epson compatible printers should use the
[Epson] option when selecting printers. Printers such
as the Panasonic KXP-1091 will work with this option.

- TI-855 printers should have the dip switch set in the
DATA-PROCESSING mode.

- Prior to printing or plotting a graph, check that the
correct device and output port have been set by
selecting SETTINGS from the CONFIGURE menu.

- When a serial port I/O error message is generated. It
is best to quit the program and restart.



The Data Analyst. ver 1.6 page 14



10. Tutorial

This section was written to guide a first-time user of the Data
Analyst through some of the more commonly used functions of the
program. The user is assumed to have a knowledge of DOS.

Begin by first copying DA.EXE and the accompanying sample data
files (with WDA extensions) to your hard disk, disk drive, or ram
drive. Be sure that you are logged to the drive and directory
containing the above files.

1. Run the program by typing DA .

2. When the title screen is displayed, hit or Alt-F
to enter the FILE menu.

3. Use the up or down arrow keys to move the hi-lighted
area to LOAD. Hit the key to select the option.

4. You will be prompted for a file specification at the
bottom of the screen. Since you are already logged
into the proper directory, just hit the key.

5. A box containing all the file names in the current
directory will appear. Use the arrow keys to move the
hi-lighted cursor to the file DATA1.WDA (this is a
sample data file with two data sets). Hit the
key to load this file. Within a few moments the file
will be loaded and the FILE menu is again displayed.

6. Depress the right arrow key twice to select the GRAPH
menu. Use the up or down arrow keys to hi-light the
VIEW option and hit . You are now in the graphics
mode. Pay close attention to the status boxes to the
right of the screen (described in the above section).
Notice that the number '2' appears in the NDS box and
the number '1' appears in the ADS box. This means that
two data sets have been loaded in memory and that the
current Active Data Set is data set #1. As can be seen
from the graph, data set #1 (symbol "o") contains ten
points and has been automatically scaled to fit the
entire graph area. A portion of data set #2 (symbol
"+") can be seen on the graph.

7. You can now familiarize yourself with some of the
functions available in this mode. Hit the F5 key once
to hide the grid. Hit F5 again to redisplay the grid.
Use F1 to enter the help screen. When you are back in
the graphics mode, use the arrow keys to move the
cursor located near the bottom of the graph. Note that



The Data Analyst. ver 1.6 page 15



the X and Y values on the display line correspond to
the current location of the cursor as if it were a data
point on the screen. Move the cursor to a position
near the center of the graph. Hit the F3 key to enter
the Label mode. The message box should display the
message "LABEL MODE". Type in the words "Sample Label"
and then hit the key. This is how user-defined
labels are entered. Now hit F2 to preview the graph as
it would appear on hard copy. Depress any key to exit
from the preview screen. Back in the graphics mode,
hit the [C] key to erase the label you have just
entered (if you wish to do so).

8. To scale the graph to fit data set #2, first press the
[I] key. Notice the number in the ADS box has changed
from '1' to '2'. Press the [S] key and the graph will
automatically be rescaled to fit the second data set.
Note that you can repeat the procedure to scale the
graph to fit data set #1 again. Assuming that your
graph is still scaled to fit data set #2, hit the [F]
key to perform a linear regression fit on the Active
Data Set (which should be 2). The message 'Curve
fitting...' will appear briefly on the message box and
then disappear. A straight line will be drawn on the
graph. This is the least squares fit on data set #2.
To see the numerical results of the analysis, press F6.
Returning to the graphics mode, you will notice that
the regression line is gone. To redisplay the line,
press [P]. To perform the same analysis on data set
#1, press [I] to switch the ADS back to 1 (the ADS box
should show '1') and then press [F]. You can press F2
to preview the graph again from here.

9. As you may have already noticed, a linear fit on the
second data set is not very precise. A second order
polynomial fit will probably fit the data set better.
To do this, press to return to the main menu.
Hit the left arrow key once to enter the STATISTICS
menu. Press to select the REGRESSION type. A
second menu indicating the available regression types
will appear. Pick POLYNOMIAL and press . Another
menu appears. Pick 2nd ORDER and press followed
by . You should now be back in the STATISTICS
menu. Press the right arrow key once to enter the
GRAPH menu. Pick VIEW and press to enter the
graphics mode a second time. Notice now near the bottom
right of the screen is the message Poly: 2. This means
that any regression analysis performed now will use the
2nd order polynomial fit.



The Data Analyst. ver 1.6 page 16



10. To perform a regression analysis on the second data set
again, first make sure that the ADS box is displaying
the number '2'. If not, press the [I] key. Now press
the [F] key to begin the analysis. A curved line
should appear within a few moments. This is the 2nd
order fit of data set #2. Press F2 to preview what you
have done so far. Note that you can have a different
type of regression analysis performed on each data set.
The resulting regression lines can be displayed or
removed. Press any key to return to the graphics mode.

11. To display the data points in the log scale, press F4.
Note that no autoscaling is possible in the Log mode.
The starting and ending Log cycle is user-defined in
the GRAPH menu in OPTIONS. Press F4 again to return to
the cartesian mode. For semi-log plots, press the [X]
or [Y] key. In all of these different scaling modes,
the regression lines that have be computed previously
for each data set can be redisplayed via the [P] key.

12. Press to return to the main menu. Press the left
arrow key once to get to the STATISTICS menu. Select
BASIC STATS and hit . A second menu will appear.
Select DATA SET 1 and press . A window displaying
all of the basic statistical information on data set #1
will appear. Press twice to return to the
STATISTICS menu. You can repeat the above procedure to
obtain statistical data for the second data set. The
program will either issue a beep or a warning message
should the user attempt to obtain statistical
information on a nonexistent data set.

13. To exit the program, press to return to the FILE
menu. Pick QUIT, hit and press [Y].

As you become more familiar with the program, you may decide to
customize more features in your graph by using the OPTIONS
selection in the GRAPH menu. You will also find that it is a
simple process to add or remove points in a data set by using the
built-in editor in the FILE menu.

*** Starting from version 1.2 of this program, the spacebar can be used
to select regression options in the graphics mode. Therefore, in
step 9 of this tutorial, the spacebar can be used to select a second
degree polynomial fit.



The Data Analyst. ver 1.6 page 17



11. User Feedback

It is only through user response that I can improve the program in
a direction that will benefit more users. All comments and
suggestions are welcomed. An EGA/VGA version of the program is
already in the works. Please direct all correspondence to me at
the address above or send E-mail to CompuServe (My CompuServe User
ID is 75136,2754).


12. General Facts

This program has been tested on the following machines:

IBM PC, IBM AT (6 and 8 Mhz), Zenith 386, TI BusinessPro,
Zenith 159, Zenith 148, AT&T 6300, IBM PS/2 30,
CompuAdd 286, CompuAdd 386, Gateway 2000 386-25MHz.

Most of the machines were fitted with the appropriate math
co-processor. Owners of 4.77Mhz machines with no co-processor
installed will probably find the program performance to be
inadequate.

Only users of DOS 3.0 or higher can use the SHELL command in the
Data Analyst without any risk. Users of DOS 2.0 or 2.1 will cause
the program to terminate unexpectedly when they invoke this option.

Generating regression fits on the Log or semi-Log scales may or
may not produce the regression line on-screen. This is due to the
fact that the resultant regression line may have negative x or y
values. In a similar vein, toggling back and forth between Log
and non-Log scales is not possible when negative data points
exist.

Users should be aware that the actual starting location of all
user-defined labels begin at the point of the small cursor when F3
is pressed (in the graphics mode). Use the X-Y coordinate display
of the cursor position as a guide to control the starting position
of the label more precisely.

The numerical and decimal point limitations of the editor
and scaling options have been noted. Work is being done on
a version of this program which circumvents these problems.

I wish to thank Anthony J. Shelton, Tze-Sen Lim, and Dr. James G.
Vaughan, all either presently or formerly with the Mechanical
Engineering Dept. at the University of Mississippi, for beta
testing the program as well as providing ideas for improving the
user interface. Ideas for additions to versions 1.2 - 1.5 are
attributed to Dr. Nadim Aziz of Clemson University.




The Data Analyst Registration/Order Form:

+---------------------------------------------------------------+
| |
| |
| To register, send this completed form along with |
| a contribution to: |
| |
| Kin-Peng Sin |
| P.O. Box 9316 |
| University, Mississippi 38677 |
| |
| Contributions of $20 or more entitles you to receive |
| the next release of this program free. |
| |
| |
| |
| |
| __________________________________________ |
| Name |
| |
| __________________________________________ |
| |
| __________________________________________ |
| Address |
| |
| _________________ __________ ________ |
| City State Zip |
| |
| |
| |
| Program version:_______________ (Displayed on title) |
| |
| |
| |
| Comments/Suggestions: |
| |
| ______________________________________________________ |
| |
| ______________________________________________________ |
| |
| ______________________________________________________ |
| |
| ______________________________________________________ |
| |
| ______________________________________________________ |
| |
| |
| |
+---------------------------------------------------------------+