R help - Nov 2011 - Replacing values in matrix/dataframe according to changing criteria

Hi there,

I am doing some network analysis working with k-cliques and over time I
want to see what nodes are members of what cliques and how big these
cliques are. I have managed to produce a matrix which shows which k-cliques
each node is part of over the 100 time periods (slow though) but I cannot
seem to calculate the size of each k-clique which is actually just a count.

Basically I have a dataframe like this with V1 being the node ID's and
V2-V4 showing which k-cliques the nodes are part of :




df <-
as.data.frame(cbind(c(1,2,3,4,5,6,7,8),c(1,1,1,1,2,2,2,2),c(0,0,0,2,0,0,0,0),c(0,0,0,0,0,0,0,0)))

# What I want is a dataframe like this where each k-clique value is
replaced by its size:

wanted <-
as.data.frame(cbind(c(1,2,3,4,5,6,7,8),c(4,4,4,4,5,5,5,5),c(0,0,0,5,0,0,0,0),c(0,0,0,0,0,0,0,0)))




It seems simple but I cannot get it working since these dataframes grow for
each time period and the k-cliques change. I have tried using a loop that
references to a table of the values but it does not work. I am sure there
is an easy way, however.



Hope to get some help,


Kind regards,


Frederik

	[[alternative HTML version deleted]]

This is a little ugly, but I think it should work pretty robustly:

T <- table(unlist(df[,-1])) # Take a close look at this to see how it
works -- it's the key to the whole thing and basically creates
something we will roughly use like a hash-table (if that term is
familiar to you) (or more accurately, like a python dictionary)

T[names(T) == "0"] <- 0 # Set zero back to zero if desired

df[,-1] <- T[match(unlist(df[,-1]), names(T))]

A heads up -- if your data are floating point numbers rather than
integers, you might run into FAQ 7.31 trouble -- that's a much tougher
question to work around, but hopefully still possible

Michael

On Wed, Nov 16, 2011 at 4:25 AM, Frederik Lyngsaa Lang
<frederiklang at gmail.com> wrote:
> Hi there,
>
> I am doing some network analysis working with k-cliques and over time I
> want to see what nodes are members of what cliques and how big these
> cliques are. I have managed to produce a matrix which shows which k-cliques
> each node is part of over the 100 time periods (slow though) but I cannot
> seem to calculate the size of each k-clique which is actually just a count.
>
> Basically I have a dataframe like this with V1 being the node ID's and
> V2-V4 showing which k-cliques the nodes are part of :
>
>
>
>
> df <-
>
as.data.frame(cbind(c(1,2,3,4,5,6,7,8),c(1,1,1,1,2,2,2,2),c(0,0,0,2,0,0,0,0),c(0,0,0,0,0,0,0,0)))
>
> # What I want is a dataframe like this where each k-clique value is
> replaced by its size:
>
> wanted <-
>
as.data.frame(cbind(c(1,2,3,4,5,6,7,8),c(4,4,4,4,5,5,5,5),c(0,0,0,5,0,0,0,0),c(0,0,0,0,0,0,0,0)))
>
>
>
>
> It seems simple but I cannot get it working since these dataframes grow for
> each time period and the k-cliques change. I have tried using a loop that
> references to a table of the values but it does not work. I am sure there
> is an easy way, however.
>
>
>
> Hope to get some help,
>
>
> Kind regards,
>
>
> Frederik
>
> ? ? ? ?[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>