Re: Xgraph Manual?

[Lloyd Wood <L.Wood@surrey.ac.uk>]

On Tue, 23 Nov 1999, K Sun wrote:

> Hello,
>     I am interested in Xgraph, but I do not know the relation between 
> Xgraph and the outcomes of ns (for example, is there any connection 
> between Xgraph and the trace file of ns?)

You parse the tracefile however you wish using awk to create xgraph
files that xgraph can draw. There are example awk scripts in
subdirectories of the ~ns/tcl directory.

You don't _have_ to use xgraph or awk; you could, say (okay, with
slightly more work) parse the tracefiles with a perl script which
wrote out a tab-delimited file you then import into something else
(such as Excel, although Excel's a toy as far as long traces are
concerned, since it imposes arbitrary limits), and the case for
replacing xgraph with something more widely-available and supported
(albeit less familiar to the ns community)[*] has been elegantly made
recently.[+]

>     I have read the examples which are attached in Xgraph so that I  
> just have a little impression how to draw a simple graph with those  
> commands. However, is it possible to automatically produce a xgraph  
> file and xgraph can give a direct display of the simuation outcome
> of ns?

Yes. Your simulation script execs something that awks the tracefile to
produce the xgraph file, and then run xgraph on that file. The test
suites can do this if you set quiet to false, so you've got lots of
example code there.

btw on xgraph: http://jean-luc.ncsa.uiuc.edu/Codes/xgraph/ appears to
have lost its webserver. This is what happens with unsupported stuff.
Hope someone has a backup copy...

L.

[*] which is the argument for incr Tcl versus OTcl, come to think of it.
[+] Guillermo Rodriguez Garcia, in
  http://www-mash.cs.berkeley.edu/dist/archive/ns-users/9911/0263.html


>     I have read the examples which are attached in Xgraph so that I 
> just have a little impression how to draw a simple graph with those 
> commands. However, is it possible to automatically produce a xgraph 
> file and xgraph can give a direct display of the simuation outcome of 
> ns?
>      In the mail archive, there is one in which Xgraph Manual has been 
> mentioned. Does it means the xgraph.man or another Doc? If it is 
> another Doc, does it has the answer of my question?
> 
>      Thanks in advance!
> 
>       Sun Kai

<L.Wood@surrey.ac.uk>PGP<http://www.ee.surrey.ac.uk/Personal/L.Wood/>