Dataplot GUI: Design and Structure

Tcl was developed by John Osterhout to add scripting capabilities to existing programs. Tk was a later extension to Tcl that allowed graphical interfaces to be developed using a scripting language (as oppossed to using a compiled language such as C/C++). Expect is a utility developed by Don Libes of NIST.

The graphical interface is written using Tcl/Tk scripts. Expect, if available on the given platform, is used to handle the communications between Dataplot and the Tcl/Tk scripts. If Expect is not available, as is the case for Windows platforms, an alternative method of communication is used that relies soley on Tcl/Tk.

See the Extending Dataplot web page for guidance in modifying the contents of these ASCII files.

1. The Tcl/Tk and Expect utilities are both freely downloadable on both Unix and Windows platforms. Maintaining Dataplot as free software that is available on a wide variety of computing platforms is a major design goal of Dataplot.

2. Tcl/Tk provides portability between Unix and PC platforms. Dataplot uses the same Tcl/Tk scripts for both the Unix and PC Windows platforms. This is significant advantage for a non-commerical program like Dataplot. We are not a software company, and wo we simply do not have the resources to develop and maintain native mode GUI's on multiple platforms.

3. Using a scripting language, as oppossed to writing C/C++ code, provided for faster development and modification of the GUI design.

4. The Dataplot GUI Tcl/TK scripts are fundamentally independent of the Dataplot code. For the original implementation of the Dataplot GUI, we made only one relatively minor change to the Dataplot source code in order to implement the graphical interface. We have since added a few minor code updates to Dataplot to implement certain desirable features to the GUI. The main point here was that we were able to add significant developments to the GUI without modifying the underlying Dataplot source code.

   This also allows us to develop new capabilities in the underlying command line version of Dataplot without dealing with the complications of incorporating them into the graphical interface.

1. Tcl/Tk and Expect have to be installed separately. This shoule be a minor issue on most platforms. The installation of Tcl/Tk is fairly automatic on the PC Windows platform and many Unix sites already provide these utilities. If your Unix site does not, you may need assistance from your local system adminstrators to have them installed.

   Some care needs to be taken that an appropriate version of Tcl/Tk is installed. Currently, we recommend at least version 8.3 of Tcl/Tk be installed (version 8.0 is the minimum).

2. As a scripting language, there may be some performance loss compared to using built-in C or C++ code. For current hardware, the Tcl/Tk performance is actually quite good. The primary area where performance is an issue is the updating of the data spreadsheet for large data sets. For very old hardware, you may wish to stick with the command line version.

3. The decoupling of the GUI from the underlying Dataplot code has significant advantages. However it does have a few drawbacks as well. Specifically, error checking of menu inputs occurs when the Dataplot commands are actually implemented, not when the items are entered into the menu.
4. Using Tcl/Tk provides a consistent interface between the Unix and Windows implementations of the Dataplot GUI. The disadvantage of this is that the Dataplot GUI will not have a "native mode look and feel" on certain platforms. This is less of an issue on Unix platforms as there is less standardization of GUI's under Unix. It is an issue on Windows platforms, where providing a "Microsoft look and feel" can make software easier to learn and use for those who have experience with other popular Windows software. Related to this on Windows platforms is the issue of windows programming. The use of Tcl/Tk requires that Dataplot be built in "console" mode. This makes certain windows programming constructs more difficult (e.g., using the windows printer drivers).

These windows can typically be controlled using standard mouse operations on your local window system. That is, they can be moved, resized, closed, and so forth. Since this involves standard mouse operations that are platform dependent (and independent of Dataplot), we don't discuss that issue further.

1. Main Menu Window
2. Alphanumeric Output Window,/a>
3. Graphics Window
4. Command Window
5. Data Spreadsheet
6. Command Line

The main menu window is at the top of the gui. A menu is accessed by clicking the left mouse button on the desire menu button.

Sub-menus may be:

1. menus to additional menus;
2. explanatory text; or
3. action menus that initiate 1 or more Dataplot commands.

Clicking the right mouse button brings up the most recently accessed menus.

For convenience, the output is saved in "pages". That is, the output for each command is saved in a separate page. The arrow buttons in the upper left corner are used to navigate these pages. A counter is displayed in the upper right corner to show the current page.

In addition, the output window shows provides the following buttons:

1. The "File" button allows the current output page to be printed or saved to a file. The items under "File" are generally self-explanatory.

2. The "Commands" button provides access to a few common commands.

3. The "Help" button access the on-line Dataplot help files.

For Unix platforms, graphics are displayed in an X11 window. For PC platforms, the graphics are displayed using a Tcl/Tk window.

The graphics window provides the following buttons.

1. The "File" button provides the following options:

   Clear - erases the graphics window

   Save Postscript - saves the current graph, in Postscript format to a file name of your choice

   Save Pixmap - saves the current graph as a pixmap that can be viewed later via the View Pixmap button

   View Pixmap - views graphs created previously by the Save Pixmap button. These graphs are displayed in a new window, not the graphics window.

   Software - generate graphs using Tcl/Tk. This is

   Graphics the default, and only choice, on PC platforms. On Unix platforms, graphics are generated using X11. You can specify the Tcl/Tk graphics by clicking this button. Using Tcl/Tk allows the window to be resized and activates the options under the Draw button.

2. The "Edit" button supports the following features.

   Settings - this brings up a table of plot control features that you can set by modifying the table. Plot control features can also be set under the "PlotMod" window.

   Graph Font - allows you to select a font for the screen graphics. Only applies if Software Graphics are being used.

   Print Font - allows you to select a font for the Postscript version of the plot.

3. The "Draw" button allows certain diagrammatic graphics to be generated using the mouse. It is only only active when the Software Graphics are being used.

4. The "Help" button access the on-line Dataplot help files for the Dataplot graphics commands.

5. The "Replot" button redraws the most recent plot.

   This is convenient if you've made a number of changes to plot control features (e.g., adding a title, setting the line types) and you want to regenerate the plot.

6. The "Refresh" button is used to redraw the current graph window. This is slightly different than the "Replot" button. With "Replot", you are actually re-entering the most recent plot command. With "Refresh", you are simply restoring the current contents of the graph window. "Refresh" is typically used when the graph has been destroyed by window operations.

7. The "Print" button prints the current graph on your default printer.

8. The Counter in the upper right corner shows the coordinates of the cursor when it is in the graphics window. These are Dataplot 0 to 100 units. This can be useful for positioning text and other graphics elements.

In addition, the output window shows provides the following buttons:

1. The "File" button allows the current output page to be printed or saved to a file. The items under "File" are generally self-explanatory.

2. The "Commands" button provides access to a few common commands.

To enter data directly from the spreadsheet window, simply move the mouse to one of the boxes, click the left mouse button, and type the value. You can use the tab key to move to the next row. Click the "Apply" button to enter the data. The "Files/Data" menu is used to read data from files.

A number of buttons are provided for this window:

1. The "Set Decimals" button allows you to specify the number of decimal points used in displaying the data. By default, the data is displayed in exponential format. You can specify 0 to 10 for the number of decimal points. Clicking "-1" resets exponential format.

2. The "Apply" button is used to read any data you manually typed in the spreadsheet into Dataplot.

3. The "Replot" button redraws the most recent plot.

   This is convenient if you've made a number of changes to plot control features (e.g., adding a title, setting the line types) and you want to regenerate the plot.

4. The "Refresh" button is used to redraw the current graph window. This is slightly different than the "Replot" button. With "Replot", you are actually re-entering the most recent plot command. With "Refresh", you are simply restoring the current contents of the graph window. "Refresh" is typically used when the graph has been destroyed by window operations.

5. The "No Update" button is used to suppress the updating of the data in the spreadsheet window. This is typically done for performance reasons. Although the data window is convenient, it can hamper performance when dealing with moderate and large sized data sets. The spreadsheet is normally updated whenever a new variable is created or a DELETE is performed. If there are several thousand data points, this can take some time.

1. Only variables are displayed. Parameters, strings, and matrices are not displayed.

   The current parameters can be listed by entering the command STATUS PARAMETERS from the command line window. Likewise, the command STATUS FUNCTIONS lists all currently defined strings.

2. Dataplot does not currently automatically read common spread sheet files such as Excel or Quattro Pro files. You need to save these files as ASCII files and then read the ASCII files into Dataplot.

If Dataplot is still executing a command initiated by one of the menus, the command line window may not accept any input. The cursor will look like a circle (it is usually an arrow) if a menu command is still executing.

This window allows Dataplot to incorporate both point and click menu operations and typed commands. Both methods of operation have their advantages, so we consider a seamless intermixing of menu and command line operation quite desirable.

A few Dataplot commands don't function properly from the gui. The details are provided under the "Performance" menu.

The "Clear" button on this window erase any input that you may have already typed. The "Variables" button brings up a list of currently defined Dataplot variables.

Privacy Policy/Security Notice
Disclaimer | FOIA

NIST is an agency of the U.S. Commerce Department.