Closed kaneplusplus closed 10 years ago
Hi Mike,
For now you need to have a few objects in your global R environment (including 'topic.proportion') in order for runVis() to work.
From the help file for runVis():
Example using Newsgroup documents from http://qwone.com/~jason/20Newsgroups/
data("Newsgroupdata", package = "LDAvis")
Run the check inputs function:
z <- check.inputs(K=50, W=22524, phi=Newsgroupdata$phi, token.frequency=Newsgroupdata$token.frequency, vocab=Newsgroupdata$vocab, topic.proportion=Newsgroupdata$topic.proportion) for (i in 1:length(z)) assign(names(z)[i], z[[i]])
Set a seed to ensure k-means produces the same clusters every time you runVis()
set.seed(333) runVis()