train_model: DEPRICATED IN V0.1.1, see new example file. Run LDA on a...

In mlinegar/litMagModelling: Easily Topic Analysis of the Modernist Journals Project with LDA

Description

DEPRICATED IN V0.1.1, see new example file. Run LDA on a Mallet instance

Usage

train_model(topic.model, dataframe, n.topics, optFreq = 20, burnIn = 50,
  numRuns = 200, num.topwords = 5, datecolname = "date",
  journalVec = NA, yearRangeRule = NA)

Arguments

topic.model	topic.model usually taken from make_model, a Mallet object
dataframe	data.frame usually taken from prep_corpus, a data.frame with fields for text, journal, date
n.topics	numeric number of topics that we want from LDA
optFreq	numeric how often should mallet optimize parameters?
burnIn	numeric number of burn in iterations for LDA
numRuns	numeric total number of runs Mallet should run before finalizing a model
num.topwords	numeric number of words to display on topic labels
datecolname	string the name of the date column from dataframe
journalVec	vector a vector of the names of journals to be included. Must be identical to values that generated topic.model possibile examples are "Blast", "Egoist", "Poetry Magazine", "Freewoman", "NewFreewoman"
yearRangeRule	string the rule specifying which years are to be included. Can take values like "> 1900", "==1919", etc. Must be identical to values that generated topic.model

Value

a data.frame of topic frequencies by document that we can analyze and graph

See Also

prep_corpus, which creates the data.frame this function draws on make_model, which creates the mallet object this function uses

Examples

date_vec <- date_vec
dataframe <- stripped_xmlData
marsden_journals <- make_model(dataframe, 10, journalVec = c("Blast", "Egoist", "Freewoman", "NewFreewoman"))
marz.df <- train_model(marsden_journals, dataframe, c("Blast", "Egoist", "Freewoman", "NewFreewoman"))

mlinegar/litMagModelling documentation built on May 23, 2019, 2:12 a.m.