Bug 476 - make sure RNG seeds and run number show up on ConfigStore output
make sure RNG seeds and run number show up on ConfigStore output
Status: RESOLVED DUPLICATE of bug 184
Product: ns-3
Classification: Unclassified
Component: core
ns-3-dev
All All
: P3 normal
Assigned To: ns-bugs
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2009-01-20 00:42 EST by Tom Henderson
Modified: 2009-11-24 00:12 EST (History)
1 user (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Tom Henderson 2009-01-20 00:42:02 EST
reminder-- the point of ConfigStore is to dump all of the pertinent simulation attributes in use for a given simulation run, and RNG seed and run number are two of the most important parameters to log, for reproducibility.
Comment 1 Mathieu Lacage 2009-01-20 08:07:45 EST
The real issue here is that the outout pf GtkConfigStore+ConfigStore is a text file which contains the attributes of all instanciated objects in a simulation but:
1) there is no object which contains the global seed+run number used
2) if there was such an object, we would have to hook an instance of it artifically in the global attribute namespace

Another option would be to extend the text file format to include also the default value of every object attribute and all global values.
Comment 2 Tom Henderson 2009-01-20 09:04:52 EST
(In reply to comment #1)
> The real issue here is that the outout pf GtkConfigStore+ConfigStore is a text
> file which contains the attributes of all instanciated objects in a simulation
> but:
> 1) there is no object which contains the global seed+run number used

agree, but I thought you might someday add support for defaults and globals

> 2) if there was such an object, we would have to hook an instance of it
> artifically in the global attribute namespace
> 
> Another option would be to extend the text file format to include also the
> default value of every object attribute and all global values.
> 

I thought this was part of the eventual plan to finish ConfigStore.
Comment 3 Mathieu Lacage 2009-01-20 09:23:12 EST
(In reply to comment #2)
> (In reply to comment #1)
> > The real issue here is that the outout pf GtkConfigStore+ConfigStore is a text
> > file which contains the attributes of all instanciated objects in a simulation
> > but:
> > 1) there is no object which contains the global seed+run number used
> 
> agree, but I thought you might someday add support for defaults and globals
> 
> > 2) if there was such an object, we would have to hook an instance of it
> > artifically in the global attribute namespace
> > 
> > Another option would be to extend the text file format to include also the
> > default value of every object attribute and all global values.
> > 
> 
> I thought this was part of the eventual plan to finish ConfigStore.


we should change the title of this bug report to reflect this then.

Comment 4 Mathieu Lacage 2009-11-23 09:23:57 EST
I believe that we do have a global for the run number and the rng seed. We just need to add support for globals on config stores

related to bug 184: finish Gtk/ConfigStore
Comment 5 Tom Henderson 2009-11-24 00:12:45 EST
(In reply to comment #4)
> I believe that we do have a global for the run number and the rng seed. We just
> need to add support for globals on config stores
> 
> related to bug 184: finish Gtk/ConfigStore

*** This bug has been marked as a duplicate of bug 184 ***