[Rsiena-help] A question on the fix (xx effects) - outdegree (density)

[Ruth M. Ripley]

Dear Joon Woo Sohn,

The degree effect will not appear if all the changes in your network are 
upwards or downwards.

The error message probably indicates a badly fitting model, as no changes 
were being made but the simulation did not end. It would not have updated 
the return object (jws.test in your example), so you would still have the 
previous entries in that object.

Please would you run a print01Report() on your data and send me the result 
direct and I will investigate.

Please also run the following command in R:

sessionInfo()

and send the result of that too, so I know what version of RSiena, R and 
what operating system you are using.

Regards,

Ruth


On Tue, 31 May 2011, Joon Woo Sohn wrote:

> Dear R-Siena Experts
>
> Hi.
>
> I am currently running a simulation to see the effect of Value homophily
> on Social Network.
>
> While running the simulation, I encountered some obstacles.
>
> 1) I have a question on the *fix(xx effects)* commands. ( there are some
> inconsistencies which I do not understand.)
> - When I use this command, sometimes outdegree (density) appears on the
> table but, sometimes it does not.
> - Also, with the same commands, sometimes the simulation works but sometimes
> it doesn't with the following message.
>
>> jws.test <-
> siena07(test.siena.model,data=sVsS.siena.data,effects=jws.test.effects,batch=F,verbose=F)
>                  sVsS.siena.objEffects.107
> "sVsS.siena|outdegree (density)|eval|Group1"
> Error in initializeFRAN(z, x, data, effects, prevAns, initC, profileData =
> profileData,  :
> invalid effect requested: see above
>
> I was wondering whether the inconsistencies are related to the occasional
> errors I get.
>
> I let one of the unresponsive runs that included density and reciprocity as
> the only effects go on, and it finished. At the end, it listed an error
> along the lines of ?unlikely to complete in this epoch.? Nonetheless, it did
> have all 10 simulated networks.
> After it finished, I couldn?t get density to show up again to test the
> effects.
>
> How would I resolve this issue?
>
> Your help is highly appreciated.
>
> Thank you.
>
> Best
>
> J
>