Description Usage Arguments Examples
Create a list of required objects from fitted topic model to visualize using d3
1 2 3 |
text |
A character vector of all training documents used to fit the LDA model |
doc.id |
An integer vector of the document ID numbers for each token occurrence in the data |
word.id |
An integer vector of the token ID numbers for each token occurrence in the data |
topic.id |
An integer vector of the topic ID numbers for each token occurrence in the data (from the fitted topic model) |
vocab |
A character vector of the tokens in the vocabulary |
K |
The number of topics in the fitted topic model |
k.clusters |
The number of clusters into which to group the topics |
lambda |
A number in [0,1] to govern the weighted average that defines a given token's relevance for a given topic. Default to 0.5. |
n.terms |
The number of terms to display on the right panel of the interactive visualization for each topic |
n.docs |
The number of example documents to display for each topic |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 | data(APinput, package="ldatools")
names(APinput) # [1] "word.id" "doc.id" "vocab" "category"
data(APtopics, package="ldatools")
names(APtopics) # [1] "topics" "loglik"
data(APcorpus)
length(APcorpus) # [1] 2250
# select only the documents that were used in training the model (i.e. remove a few very short, or empty docs)
text <- APcorpus[APinput$category == 0]
# Run the function and write the JSON file:
z <- jsviz(text=text, doc.id=APinput$doc.id, word.id=APinput$word.id, topic.id=APtopics$topics, vocab=APinput$vocab,
K=30, k.clusters=1, lambda=0.5, n.terms=30, n.docs=10)
# Write the list to a JSON object and place in a directory from which to serve the d3 webpage:
library(RJSONIO)
z.out <- toJSON(z)
cat(z.out, file="path/lda.json")
# Now serve index.html from path/
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.