[Rsiena-help] help with composition change file

[Ruth Ripley]

Dear Anne,

You need to use the function sienaCompositionChangefromFile to read your 
composition change data in. The alternative list format is rather 
difficult to produce. as.list() on a data frame creates a list of 
columns not rows.

Regards,

Ruth

On 06/12/2011 19:23, ter Wal, Anne wrote:
> Hi,
>
> I am currently running SIENA models in R, looking at the impact of
> geographical distance (as a dyadic covariate) in network evolution.
>
> I have ran models with two waves successfully by analysing the evolution
> among two network observations with equal node sets.
>
> Now I am trying to analyse five waves of data (with geographical
> distance - recoded as integers between 0 and 255 - as a changing dyadic
> covariate).
>
> This requires, however, to use a file that indicates the composition
> change. As actors only exit and enter once, I have made a text file with
> two columns for each actor in the network, where numbers between 1 and 5
> indicate the time of entry and time of exit (separated by a tab).
>
> My syntax is as follows:
>
> # load libraries
>
> library(snow)
>
> library(network)
>
> library(rlecuyer)
>
> library(xtable)
>
> # set directory
>
> Setwd("….")
>
> list.files()
>
> # load network data
>
> network.data.w1 <- as.matrix(read.table("Network 1978-1982-b-Matched.txt"))
>
> network.data.w2 <- as.matrix(read.table("Network 1983-1987-b-Matched.txt"))
>
> network.data.w3 <- as.matrix(read.table("Network 1988-1992-b-Matched.txt"))
>
> network.data.w4 <- as.matrix(read.table("Network 1993-1997-b-Matched.txt"))
>
> network.data.w5 <- as.matrix(read.table("Network 1998-2002-a-Matched.txt"))
>
> distance.data.w1 <- data.matrix(read.table("Geomatrix_1982_all_int.txt"))
>
> distance.data.w2 <- data.matrix(read.table("Geomatrix_1987_all_int.txt"))
>
> distance.data.w3 <- data.matrix(read.table("Geomatrix_1992_all_int.txt"))
>
> distance.data.w4 <- data.matrix(read.table("Geomatrix_1997_all_int.txt"))
>
> composition.change.data <- as.list(read.table("composition_change.txt"))
>
> # a. set network dependent variables
>
> coinv <- sienaNet(array(c(network.data.w1, network.data.w2,
> network.data.w3, network.data.w4, network.data.w5), dim=c(425, 425, 5)))
>
> # b. set attributes
>
> distance <- varDyadCovar(array(c(distance.data.w1, distance.data.w2,
> distance.data.w3, distance.data.w4), dim=c(425,425,4)))
>
> compChange <-sienaCompositionChange(composition.change.data, option=1)
>
> # c. data specification
>
> mod1 <- sienaDataCreate(coinv, distance, compChange)
>
> But I get the following error message:
>
> Error in sienaDataCreate(coinv, distance, compChange) :
>
> composition change incorrect size: compChange
>
> Do you have any idea what is going wrong here?
>
> I also tried to import the file "composition_change.txt" as a matrix,
> but then I got another error earlier, namely at the command
> sienaCompositionChange
>
> Error in sienaCompositionChange(composition.change.data, option = 1) :
>
> changelist must be a list
>
> I hope you can help me in resolving this problem.
>
> Thanks a lot for your help in advance.
>
> Kind regards,
>
> Anne ter Wal
>
> *Dr. Anne ter Wal*
>
> /Research Fellow/
>
> Imperial College Business School
>
> Innovation and Entrepreneurship Group
>
> Tanaka Building, South Kensington Campus
>
> London SW7 2AZ
>
> E: a.terwal at imperial.ac.uk <mailto:a.terwal at imperial.ac.uk>
>
> W: www.anneterwal.com <http://www.anneterwal.com/>
>
>
>
> _______________________________________________
> Rsiena-help mailing list
> Rsiena-help at lists.r-forge.r-project.org
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rsiena-help