Downloading DATAPLOT for Mac OS X
|
|
Contact Information
|
February 2008.
Alan Heckert
Statistical Engineering Division
Room 331, Building 820
National Institute Of Standards and Technology
Gaithersburg, Md. 20899-8980
(301) 975-2899 or FTS 879-2899
alan.heckert@nist.gov
|
Jim Filliben
Statistical Engineering Division
Room 353, Building 820
National Institute Of Standards and Technology
Gaithersburg, Md. 20899-8980
(301) 975-2855 or FTS 879-2855
james.filliben@nist.gov
|
|
|
Introduction
|
This page contains the intstructions for downloading and
installing Dataplot on the Mac OS X operating system.
My Mac OS X platform is a Power Mac (i.e., non-Intel) running
Mac OS X 10.4 (tiger). If you have a compatible system, you
can install a pre-built Dataplot executable (i.e., a binary
install). If not, you need to download the source (i.e.,
a source install).
Be aware that this is a "Unix/X11" port as oppossed to a
"native" Mac OS X application. There are currently two
ways you can run Dataplot under Mac OS X.
- You can run Dataplot in a traditional command line
mode (this will be referred to as the command line
version).
Note that the command line version can be run from
either a standard Mac OS X terminal window or an
X11 terminal window.
- Alternatively, you can run a Graphical User Interface
(GUI). The GUI for Dataplot is written using the
Tcl/Tk scripting language (we will say more about
this below).
|
|
Prerequisites
|
|
g77, X11, Aquaterm, and Tcl/Tk
|
Dataplot requires several external programs be installed.
Specifically,
- g77 - even if you install the pre-built Dataplot
executable, you need the run time libraries for
the g77 compiler.
If you do not already have g77 installed, you can
download the g77 installation file. The install
script described below will perform the installation.
- X11 - Dataplot requires that X11 be installed on
your Mac OS X system.
The Apple web site provides further information for
installing and running X11 under Mac OS X. This page
also contains a good introduction to using X11 on
Mac OS X.
- Aquaterm - The February 2008 version of Dataplot added
an Aqua
driver for screen graphics. This means you can now
run the command line version of Dataplot in a standard
terminal window (previously you had to run in an X11
window if you wanted screen graphics).
Dataplot implements the Aqua driver using the
Aquaterm library. You can download Aquaterm from the
Aquaterm web site. For convenience, you can also
download it
directly from this page.
- Tcl/Tk - As mentioned above, the Dataplot GUI is
implemented using the Tcl/Tk scripting language.
As there are other applications that use Tcl/Tk, you
may want to check if to see if it is already installed.
Look for either "/usr/local/bin/wish" or "/usr/bin/wish".
If one of these exists, then Tcl/Tk should already be
installed.
Tcl/Tk is an open source scripting language that is
supported on multiple platforms. In particular, there
are two primary sources for downloading and installing
Tcl/Tk for Mac OS X:
- Tcl/Tk Aqua Batteries-Included can be
downloaded from the SourceForge web site.
For convenience, you can also download
it directly from this page.
Note that this is the version I am currently
running on my own Mac. The Source Forge web
site contains several versions of the download.
The Dataplot Tcl/Tk scripts currently only use
the core Tcl/Tk (i.e., no packages).
- The
ActiveState web site also has a version
of Tcl/Tk that has been optimized for the
Mac OS X platform.
Note that although this implementation is freely
downloadable, you have to download it directly
from the Active States web site.
You should be able to run Dataplot with either of these
implementations. However, be aware that they install
Tcl/Tk in a different directory.
|
|
Downloading and Installing Dataplot Files
|
|
|
Assuming that the prerequisite software has been installed,
you can proceed with the downloading and installation of the
Dataplot files.
I have modified the installation a bit starting with
the February 2008 version of Dataplot. The
older installations will
still be available for the near future.
|
|
Step 1: Binary or Source Installation
|
As mentioned above, you can either perform a binary installation
or a source installation.
I develop for the Mac OS X using a Power Mac running tiger
(10.4). So if you have an Intel-based Mac or you are running
10.2 or earlier, you will probably need to do a source
installation. Otherwise, you should be able to use the
binary installation.
Note: If there are any users who have an Intel-based Mac and
would like to contribute the binaries, please contact
alan.heckert@nist.gov.
|
|
Step 2: Download the Files
|
The next step is to download the files. You can download
these files wherever it is convenient, but you should download
them all to the same directory.
- The file dpbin.tar.gz
contains:
- The Dataplot executable for the Power Mac
(installed in /usr/local/bin).
- A script to run the Dataplot GUI
(installed in /usr/local/bin).
- Dataplot uses the GD library to generate JPEG, PNG,
and GIF format files. The GD library in turn uses
the PNG library, the ZLIB compression library, the
JPEG library, and the Freetype library. Dataplot
will install the versions of the libraries used by
Datplot in the "/usr/local/lib" directory.
The current version of the Mac OS X Dataplot
executable uses version 2.0.33 of GD, version 1.2.8
of PNG, version 1.2.3 of ZLIB, and version 2.10 of
Freetype.
If you are going to perform a source build, then you
do not need to download this file.
- The file dplib.tar.gz
contains the Dataplot auxillary files. This includes the
online help files, sample data files, the script and
menu files for the GUIs, and other files used by
Dataplot. These files are all installed in the
"/usr/local/lib/dataplot" directory.
Note that this file should be downloaded whether you
performing the binary install or the source install.
- The file dpsrc.tar.gz
contains the Dataplot source code files. This includes
Fortran source files, C source files, include files,
and a script file to build Dataplot for Mac OS X.
These files are all installed in the
"/usr/local/src/dataplot" directory.
Note that this file does not need to be downloaded
if you are performing the binary install.
- The file
install_dataplot_files contains a script for
unpacking the files.
|
|
Step 3: Unpack the Files
|
The next step is to run the "install_dataplot_files" script.
This script will unzip and then unpack the downloaded files to
appropriate locations.
You should edit the file using any ASCII text editor and
make the following changes (if needed):
- If you need to install the g77 compiler, look for the
lines
# gunzip g77.tar.gz
# tar -P -xvf g77.tar
and delete the leading "#" to uncomment these lines.
- If you are performing a source install, look for the
lines
gunzip dpbin.tar.gz
tar -P -xvf dpbin.tar
and add a "#" at the beginning of the line to make these
lines comment statements.
- If you are performing a source install, look for the
lines
# gunzip dpsrc.tar.gz
# tar -P -xvf dpsrc.tar
and delete the leading "#" to uncomment these lines.
Then open a terminal window, move to the directory containing
the installation files, and enter the command
sudo ./install_dataplot_files
This will unpack the files and install them in the appropriate
directories. The "sudo" command gives you temporary
administrator privleges (you need this since the Dataplot
files will be installed in "/usr/local").
|
|
Binary Installation
|
|
Step 1: Check Path
|
The Dataplot executables are installed in "/usr/local/bin"
directory. This means that "/usr/local/bin" should
be in your default path. To check this, enter (from a
terminal window)
If Dataplot comes up, then your path is ok. Enter QUIT
to exit Dataplot. On the other hand, if you get a
"dataplot not found" message, you need to update your
default path.
If you use the c-shell or the t-shell, then you need to add
a line like the following to your .cshrc file:
set path=(/usr/local/bin $path)
If you use the Bourne or Korn shell, then you need to add the
lines like the following to your .bash_profile file:
PATH=$PATH:/usr/local/bin
export path
To activate this path, close your current terminal window
and then re-open it. Now when you enter
it should come up correctly.
|
|
Test the Command Line Executable
|
The command line version of Dataplot can run in either
a standard Aqua terminal window or in an X11 terminal
window.
To test the command line version in a standard Aqua terminal
window, enter the following commands
dataplot
device 1 aqua
call minitest.dp
quit
To test the command line version for X11, open an xterm window
and enter the following commands
dataplot
device 1 x11
call minitest.dp
quit
|
|
Test the GUI Version
|
Before testing the Dataplot GUI, you should first check
the following:
If you do need to modify either of these files, you need
to edit using the "sudo" command since these files are in
the "/usr/local" directory. For example,
sudo vi /usr/local/bin/xdataplot
If you installed the Source Forge Tcl/Tk package, you should
intitiate the Dataplot GUI from an xterm (X11) terminal window.
If you installed the Active State Tcl/Tk package, you can
initiate the Dataplot GUI from either an xterm terminal
window or from a standard Aqua terminal window.
To test the Dataplot GUI version, open an appropriate
terminal window and enter the command
This should bring up the GUI. You should see six windows.
- A top level window containing the menus (e.g.,
Files/Data, Plot).
- The upper left window is where Dataplot
alphanumeric output is generated.
- The upper right window ("Graph") is where
Dataplot graphs are generated.
- The lower left window stores the Dataplot commands
that have been issued.
- The lower right window is a data spreadsheet.
- The window labeled "Command Line" allows you to
enter Dataplot commands directly.
Click the "Exit" button on the top right menu window to
close the GUI.
If the Dataplot GUI does not come up correctly, the problem
is usually that the paths to the Tcl/Tk software are not
set correctly.
|
|
Source Installation
|
|
Building from Source
|
If you are unable to use the pre-built Dataplot executable,
then you can build Dataplot from the source code.
The installation script above should have installed the
g77 compiler (the gcc compiler should already be installed).
Although I haven't yet automated this to the degree that I
ultimately hope to, it should be a relatively straightforward
source build if you skip the GD library.
|
|
You Can Move the Files to a More Convenient Directory
|
Although Dataplot will install the source files to
"/usr/local/src/dataplot", you will probably find it more
convenient to move the files to a directory in your own
file space so you won't have to use the "sudo" command to
modify files. For example, you might want to create the
directory "~/src/dataplot" and then copy the files from
/usr/local/src/dataplot to ~/src/dataplot.
There is no particular restriction or structure required
for the source directory.
|
|
GD (and associated libraries)
|
As mentioned above, Dataplot uses the GD library to create
graphics in JPEG, PNG, and GIF format. Although this is
certaintly a useful feature, it is not critical for running
Dataplot.
If you are uncomfortable in building software from source,
then I recommend building Dataplot without the GD
library. Even if you are more experienced with this, I
recommend first building Dataplot without the GD libraries.
Once the basic Dataplot is working, you can go back and try
to build it with the GD library enabled.
In any event, when you want to enable the GD library, you
will need to install the following libraries:
- GD
- Freetype
- JPEG
- PNG
- ZLIB
First, check to see if these libraries are already installed
in either "/usr/lib" or "/usr/local/lib". The ZLIB library
(look for files starting with "libz") will typically be
installed. The others may or may not be.
For convenience, I have provided links to these libraries
Note that you should build the ZLIB library first, then
the PNG, JPEG, and Freetype libraries, and finally the
GD library.
Each of these comes with its own installation instructions.
|
|
Edit the Build Script
|
The source directory should contain the file
Edit this file using an ASCII editor.
- Near the beginning of the file should be the line
that starts with "cd ". Edit this line to move
to the directory where you have the source files.
- Look for the line "set GD_FLAG = on". To disable the
GD library, change the "on" to "off". If you have
successfully installed the GD library, you can
leave this as "on".
- Look for the line (at the end of the file) that starts
with "$FCOMP -dynamic -o dataplot_macosx main.f". The
file after the "-o" specifies the name for the Dataplot
executable. You can change this to whatever you like.
No other edits should be required. To initiate the build,
enter the command
Once you are satisfied that you have a successful build, you
can move the executable to "/usr/local/bin".
|
|
Problems with Downloading
|
|
Installation Available on CD
|
Sometimes users do have problems downloading Dataplot from
the web pages. Note that I do not maintain the NIST ftp/web
servers (I do provide the content for Dataplot, but I am
not a system adminstrator for these servers). Downloading
problems may be caused by firewalls, ftp/browser
configurations, etc.
I simply do not have the time or expertise to diagnose these
types of problems. I can check to see if our ftp server is
up, but not much beyond that.
If you are unable to download via the web, you can do the
following:
- You can try using anonymous ftp directly.
ftp ftp.nist.gov
user name> anonymous
password> your e-mail address
>cd /pub/dataplot/mac
>dir
At this point, the dir command will list the files
needed for the Mac OS X installation. Be sure to enter
a "binary" command before downloading any "exe",
"tar.gz" or "tar.Z" files.
I only recommend this if you already reasonably
familiar with using anonymous ftp.
- Alternatively, I can send the Dataplot installation on
a CD-R. This is easier for me than trying to
diagnose ftp/browser/firewall problems. There is no
charge, just send a request to
Alan Heckert.
Be sure to include your mail address (not e-mail) and
the desired platform.
|
|
Installation Problems
|
If you have problems with the installation, contact
Alan Heckert
for assistance.
|
|
Date created: 6/5/2001
Last updated: 2/6/2008
Please email comments on this WWW page to
sedwww@cam.nist.gov.
|
|