Simulates a Multinomial LDA.
# S3 method for rlda
generateMultinomialLDA(seed0, community, variables,
observations, totalElements, beta, gamma, ...)Initial seed to simulate a Multinomial LDA.
Total number of latent clusters. Must be greater than 2.
Total number of variables. Must be greater than the number of communities.
Total number of observations. Must be greater than 1.
Total expected number of elements for each observation. Must be greater than 1.
Vector of positive hyperparameters with dimension equal variables.
Scalar hyperparameters that must be positive.
other arguments may be useful.
Generates a list with the simulated Theta and Phi matrix of parameters, Z latent matrix of communities and and Data matrix for the Multinomial LDA.
# NOT RUN {
#Generate fake data
res<- generateMultinomialLDA.rlda(seed0=5587, community=4,
variables=100, observations=1000, totalElements=40,
beta=rep(1,100), gamma=0.01)
#Show results
res
# }
Run the code above in your browser using DataLab