*****************************************************
* *
* TechPlot *
* Technical Plotting and Data Processing *
* TEST DRIVE v1.01 *
* *
* (c) Polysoft 1992 All Rights Reserved *
* *
* You may make copies or pass it to your *
* friends freely *
* *
* Readme.doc *
* *
* Thank you for downloading TechPlot *
*****************************************************
Contents
Part I. Getting Started
1. Starting and exiting
2. Accessing Help
3. How to contact Polysoft
Part II. Tutorial
Lesson 1: Data scatter plot and smooth curve
Lesson 2: Multiple plot types
Lesson 3: 3D plots
Lesson 4: Curve fitting, confidence and prediction intervals
*************************************************
Part I. Getting Started
*************************************************
1. Starting and exiting
Once you have installed TechPlot , you should change to the directory
that contains the TechPlot program files and type TECHPLOT at the DOS prompt.
When you finish working with TechPlot, choose the File | Exit command
or press . If you have made any changes you haven't saved, TechPlot
will ask if you want to save those changes before quitting.
2. Accessing Help
You can get Help in TechPlot in the following two ways:
ú Press or select Help on the main menu to bring up the Help menu,
and then choose a proper Help command.
ú Press to display the help index directly, without going through the
Help menu.
3. How to contact Polysoft
If you should encounter any problems with TechPlot, or if you have any
other comments or suggestions, please send your mail to
Polysoft
P.O. Box 526368
Salt Lake City, UT 84152
You can also call our technical support line between 9:00 a.m. and
5:00 p.m. Mountain Time at (801) 485-0466, or fax us at (801)-485-0480.
*************************************************
Part II. Tutorial
*************************************************
Welcome to the TechPlot tutorial. The following four lessons provide
step-by-step instructions for designing a variety of plots. These lessons
illustrate how various TechPlot functions work together to create
publication-quality graphs.
Lesson 1: Data scatter plot and smooth curve
In this lesson, we will create the following plot as shown in Figure 1.
This plot consists of a scatter data and a smooth curve. The smooth curve
is created by a polynomial fitting of the observed data points.
Step 1: Loading data from a disk file
To load a data file from a disk file:
ú Choose the File | Open - Data Sheet command. The Open File dialog box
will be displayed.
ú Select the data file named LESSION2.PSD, then click Open.
This data file contains two columns of data ('TIME' and 'CONCENTRAT').
Step 2: Fitting data via polynomial
To fit the observed data via polynomial, follow the steps below:
ú Choose the Math | Fitting - Polynomial command, the Polynomial dialog
box will be displayed.
ú Change the weighting factor to '2' and the order of polynomial to '5',
then click OK. This will start the fitting process.
After the fitting is completed, the Save Result Option dialog box will be
shown on the screen to allow you select result saving options. For this
example, save only calculated result. (This is the default setting) Click OK.
Once the Save Result Option dialog box disappears, the Calculated Result
dialog box is displayed with the result of curve fitting including parameter
statistics. There is only partial of the result shown in the file viewer.
To view the entire file, use the up and down arrow keys or the and
keys. You can save this to a disk file by clicking on the Save
button, or print it to a printer by clicking on the Print button. Choose OK
to exit this dialog box.
The calculated column be attached to the data sheet (immediately to the
right of the existing columns). For this example, the column names will be
'Ind3' and 'Dep4'. 'Ind3' means it corresponds to an independent variable
and the current column number is 3. 'Dep4' means it corresponds to a
dependent variable and the current column number 4.
Step 3: Selecting plot type and options
To create a new graph, choose the Plot | 2D Curve - XY Line command. We
are going to add two curves to the plot:
ú Select 'TIME' as x-axis and 'CONCENTRAT' as y-axis for curve #1, then
click the Add Curve button or press .
ú Select 'Ind3' as x-axis and 'Dep4' as y-axis for curve #2, then click
the Add Curve button or press .
ú Click OK.
Since curve #2 is the fitted curve, it should be plotted as a continuous
curve. TechPlot will calculate extra points to make it a smooth curve. You
will see a green bar growing on the status bar which indicates the progress
of smooth curve calculations. Once the calculation is done, the plot is
created and shown in a new Plot View window.
Step 4: Changing the curve format
By default, all data are connected by lines. Since you want to compare
the observed data with fitted curve, you should make the observed data
points disconnected. To do this, you simply select curve #1 by clicking on
it and choosing the line tool (line button [\]) in the edit toolbox.
You may also disable the x-grid lines: double click the graph or select
the graph and choose the Format | Object Format command, the Frame Format
dialog box will be displayed. Uncheck the XGrid check box and click OK.
Step 5: Adding annotation and legend
To add annotations to the graph, perform the following steps:
ú Choose the Text Tool ([T] button) and click mouse at the position where
you want put the text.
ú The Text Format dialog box will be displayed immediately. By default,
the text string is 'Hello'. You can change it to whatever you want.
You can add 'Observed Data' and 'PolyFitted Curve' to the plot. Make
the size of both text strings to '30' by '30'.
After completing steps 1 - 5, you will see the Figure 1 on the screen.
Lesson 2: Multiple plot types
The objective of this lesson is to create a graph that contains multiple
plot types. We will create a graph that contains two type of plot: a smooth
curve and a histogram chart. The graph will be rotated by an angle of 35
degree. The legend of this graph are grouped objects of symbols and text
objects.
Step 1: Automatic generating data
As an example, we will generate data inside TechPlot. Perform the
following steps to generate a Gaussian distribution data set:
ú Choose the File | New - Data Sheet command to create a new data sheet.
ú Choose the Data | Fill Selection - Gaussian command to create a Gaussian
Normal distribution data column.
ú In the Fill Selection: Gaussian dialog box, we enter the data range from
column 1 to 1 and row 1 to 5000. We use 0 as mean and 1 as deviation.
Click OK.
The 5000 data points then are created and put on the data sheet. Change
the column name to RawData. After this, choose the Math | Special -
Histogram command to create histogram data. This command will display the
Create Histogram Data dialog box. In this dialog box, enter '15' for the
number of intervals, the destination of resulting columns will be '2' and
'3'. Click OK or press . The resulting column names are 'Intvl2'
and 'Count3' which stand for interval number and data number counts,
respectively.
Step 2: Creating a histogram plot
Choose the Plot | 2D Special - Histogram command to create a histogram
plot. In the plot setup dialog box, choose column 'RawData' and enter '15'
for the number of intervals in the Interval field, then click OK. Then plot
is created and shown in the Plot View window.
Step 3: Switching between Plot View and Sheet View.
Switch back to the data sheet by choosing the Window | Next () command
(or Window | List if you have more than two windows opened).
Now, we fit the 'X' and 'Y' data using a natural spline with 6 nodes. To
do this, choose the Math | Fitting - Spline command. In the Natural Spline
dialog box, set column 'Intvl2' as Independent Variable, and column 'Count3'
as Dependent Variable. We also set the node to '6' and weight factor to
'0'. Click OK to start fitting. After successful fitting, we save the
results (just the calculated result, not confidence and prediction interval).
The resulting columns are 'Ind4' and 'Dep5'. After the above math
operations, switch back to the plot view by choosing the Window | Next
() command.
Step 4: Adding another smooth curve
Now, let's add the spline fitted data to this histogram plot:
ú Select the histogram graph, not on the bars.
ú Choose the DataPlot | 2D Curve-XY Line command.
ú Add a new curve with 'Ind4' as 'X' and 'Dep5' as 'Y'.
The following plot will be shown on your computer screen.
Step 5: Rotating a graph
Select the graph and choose the View | Rotate command. An instruction
message in the status line says "use arrow keys to rotate, and Enter key to
stop." So, use the arrow keys to rotate the plot! While you are rotating
the graph, a frame box will show the current position of the plot. Once a
satisfactory position is reached, press to end the rotating action.
Next, change the x-axis range to [-4,4] with major tick mark number 7 and
minor tick number 1. Disable both x-axis and y-axis grid lines and change
the plot title to 'Multiple Plot Types'.
Step 6: Grouping objects
After rotating the graph, you can add some annotations.
First, add some descriptive text for both curves. Second, add a math
equation to describe the Gaussian distribution. To do this, choose the
text tool and click on the spot where you want to put the equation. For
this example, choose the right-top corner. Then type in the text exactly
as shown below:
Y=K*e\u-c*(X-m)\u\s2
Group all of these annotations together and move to somewhere else.
Choose the Option | Group command, then move your mouse cursor to the
left-top corner of the text. Press the mouse button and drag it to include
all text objects. A big rectangle with four highlighted corners will
surround these objects. Now these text objects are grouped as one object.
You can then select it or move it to your desired location. (to left-top
corner, for example)
After these operations, you will see the graph of Figure 2 on your screen.
Lesson 3: 3D plots
In this lesson, you will learn several 3D plot editing techniques.
Step 1: Creating a 3D surface graph
Load the data set LESSON7.PSD. This data set contains three columns of
data as listed below:
Latitude Longitude Collection
1 4 5
2 2 3
6 3 5
4 1 4
3 3 6
5 2 7
We will create a 3D surface (mesh) graph using observed data by means of
the 3D-mesh option. For this example, we will use the powerful 3D mesh
creation feature to create a 3D mesh surface with 30 meshes. To do this,
perform the following steps:
ú Choose the Plot | 3D Surface - XYZ command.
ú Enter the following dialog box data:
Latitude -> X; Longitude -> Y; Collection -> Z
xMeshNum=30, yMeshNum=30
ú Choose OK to create a new graph.
The default angle rotation for X, Y, Z axes are: 0, 30 40. To see this
graph from another angle, let's resize and rotate it. First, perform a 2D
resize action by selecting the graph and dragging one of the corners to a
bigger size such that the entire graph fills the Plot View window. Then,
do the 3D rotation as follows:
ú Select the graph and choose the View | Rotate command. Notice that
the toolbox is changed to 3D rotation tools with arrow buttons and
input-lines.
ú Change axes angles. You can either click on those arrow buttons or type
in a number in the inputline box and click the Apply button. For precise
demonstration, type in the angles for x, y, and z axes as 0, 25, 132,
respectively.
ú Click the Done button to go back to the normal Plot View.
Now the graph is better looking and is shown in Figure 3.
Step 2: Creating 3D contour plot
Make a copy of the above surface plot by selecting it and choose the
Edit | Copy command. Then create a new Plot View window by choosing the
New | Plot command, and paste the surface plot to the new Plot View window
by choosing the Edit | Paste command. In order to add contour level lines
to the 3D surface, you must select it and then choose the Format | Contour
command. The Contour dialog box is then displayed. In this dialog box, a
default value of contour level is show in the New Contour input-line. This
value is the middle value of the z-axis range. You can then enter new
values and use the Add command button to add contour level values. The
maximum number of contour lines is 10. For this example, we add seven
contour level values: 3.5, 4, 4.5, 5, 5.5, 6, 6.5. Change the x-axis range
to [1,6] and z-axis range to [0,8]. The graph is shown in Figure 4 below.
Step 3: Making a 3D function plot
Now let us create a 3D spherical surface graph. To do this, we first
create a new Plot View by choosing the File | New - Plot command. Then
choose FuncPlot | 3D Sphere Surface command. You will see the default
dialog box setting as follows:
Z=Y/360.0*cos((X-90)*PI/180).
xMin:=0.0; xMax=180; xSteps=20.
yMin=0.0; yMax=450; ySteps=30.
Click the OK button to accept the default settings. When the new graph is
created in a Plot View window, rotate it to (0, 30, 150) and change its
x-axis range to [-0.2, 1]. The new graph is shown in Figure 5.
Lesson 4: Curve fitting, confidence and prediction intervals
Nonlinear curve fitting is very useful in developing experimental models
and in data processing. TechPlot offers several least-squares fitting
methods and parameter statistics analyses. Among these features, the
confidence and prediction intervals are often used to verify the goodness
of a model and goodness-of-fit to the experimental data. In this lesson,
you will create a graph with observed data, fitted curve, confidence
intervals, and prediction intervals.
Step 1: Performing user-defined nonlinear curve fitting
Load the data set LESSON8.PSD. This data set contains two columns of data.
To fit the data using a user-defined model, we choose the Math | Fitting -
User-defined command. The Fitting dialog box is then displayed. In this
dialog box, the default model template is shown in the model editor. Next
load the model file LESSON8.EQN by choosing the Load command. The model
equation is as follows:
[INDVAR]: T
[DEPVAR]: Y
[PARAMS]: A, B, C, D
[EQUATIONS]:
Y=A/(1+EXP(B*(TT-C)))+D
[INIT PARAMS]:
A=115
B=0.12
C=110
D=42
ENDMODEL
Step 2: Changing the error in the model equation editor
Choose the Compile command to compile the model. An error message will be
displayed to indicate that there is a missing variable. This error is
reported by the built-in function parser. In the above equation, there is
a typo in the 'Y=A/(1+EXP(B*(TT-C)))+D' equation. 'TT' should be 'T', the
independent variable. To correct the error, move the mouse cursor to the
equation line that contains 'TT' and click on 'TT'. Use the
key or the ~~ key to delete the extra 'T'. Then choose the Compile
command again. If the compilation is successful, all the three numerical
method command buttons will be activated. For this lesson, we choose the
Powell's method to fit our data. That is the Powell command button.
Step 3: Saving the fitting result
After the fitting is completed, the Save Result dialog box will be
displayed. In this dialog box, you can check the saving options. For this
lesson, we save all data (data, conf, pred, parameters).
After the Save Result dialog box disappears, the Calculated Result dialog
box is shown. All the parameter statistics and goodness-of-fit statistics
are listed in the file viewer.
After calculation, all data are shown in the data sheet. The observed
data columns are 'T' and 'Y'; Calculated data columns are 'Ind4' and 'Dep5';
Parameter values in column 'Param3'; Confidence interval columns are
'ConfL7' and 'ConfH8'; Prediction interval columns are 'PredL9' and
'PredH10'.
Step 4: Creating the plot
Choose the Plot | 2D Curve - XY Line command and add observed data and
calculated curves to the same graph. To do this, perform the following
steps:
ú Add curve #1: observed data 'T' and 'Y'.
ú Add curve #2: calculated data 'Ind4' and 'Dep5'.
ú Add curve #3: lower confidence interval band 'Ind4' and 'ConfL6'.
ú Add curve #4: upper confidence interval band 'Ind4' and 'ConfH7'.
ú Add curve #5: lower prediction interval band 'Ind4' and 'PredL8'.
ú Add curve #6: upper prediction interval band 'Ind4' and 'PredH9'.
ú Click the OK button. Then the graph is shown in a new Plot View
window.
Next, create several text objects to mark each curve. Also edit the model
equation on the top-right corner with parameter values. The annotations for
each curve and the parameter values are straightforward. Creating the
equation is more complicated. Use the equation editor to create 1+e^B(T-C)
first, then create other text objects 'Y=', 'A', and '+D'. Use the editor
tools to move them together and make the equation. Finally, use the group
option to group these text objects together. After they are grouped, you
can treat them as one plot object. With some arrangement, you should be a
ble to see the following graph on the screen..
After all these lessons, you should be an expert on TechPlot. Use it and
have fun!
~~