blei-lab / lda-c Goto Github PK

Is there a diffrence between the method you mentioned in your paper 《Latent Dirichlet Allocation》and the code?
Mainly in the proccess "E-M" step.
In the paper you used variational "inference" , but in the code you used something about "sufficient statistics"?

Hi,
I am trying to estimate the perplexity on a test set (unseen set of documents). After the inference step on the test set, I see the likelihood file has large negative numbers (e.g. -1800). What are these numbers exactly. If these are log likelihood estimates, should we compute the perplexity by just taking the exponent of average of these values.

Looking forward for an answer.
Thanks,
Jagdeep

Hi，I am a newer to use lda-c and i couldn't find the function 'lgamma', and i think if this function is 'log_gamma'. thank you!

I would like to reproduce the the results in https://github.com/Blei-Lab/lda-c/blob/master/example/ap-topics.pdf however I don't know what settings.txt alpha and seedings were used. Can you please help?

I'm a beginer of LDA. pls tell me how to use this lda command.
usage : lda est [initial alpha] [k] [settings] [data] [random/seeded/*] [directory]
lda inf [settings] [model] [data] [name]

in my mind. this is a tool to get the topic words of an article.
so if I have handreds of articles by hand(like 0000.txt-1000.txt). how can I use lda to get the topic words of an article?

many people coding python
it would be very kind of you to share this code in python
thanks