#
# Help!
#

help(c)

?c

#
# Vector
#

c(2,3,5) 

c("aa", "bb", "cc", "dd", "ee")

c(TRUE, FALSE, TRUE, FALSE, FALSE)

#
# Matrix
#

m <- matrix(data = c(1,2,3,4,5,6,7,8,9,10,11,12), 
            nrow = 3, 
            ncol = 4)
m 

#
# List
#

numbers <- c(2, 3, 5) 

strings <- c("aa", "bb", "cc", "dd", "ee") 

my.list <- list(numbers, strings, 3)

my.list 

#
# Data frame
#

teams <- c("PHI","NYM","FLA","ATL","WSN")
wins <- c(92,89,94,72,59) 
losses <- c(70,73,77,90,102) 

table.data <- data.frame(teams, wins, losses) 

table.data 

#
# Basic R functions
#

c() # combine two or more elements into an object

class() # explore elements’ data class
length() # explore number of first dim. of object
dim() # explore dimensions of two-dimensional obj.
nrow() # number of rows
ncol() # number of columns

head() # first few rows of data 
tail() # last few rows of data 
str() # explore structure of object

names() # names in the named vector - one dimension 
rownames() # names of rows - two dimensions 
colnames() # names of columns - two dimensions

#
# working directory
#

setwd("C:\\Users\\Lukas\\Desktop\\data\\")

setwd("C:/Users/Lukas/Desktop/data/")

#
# libraries
#

library("tm")

require("tm")

detach("package:tm", unload = T)

tm::Corpus()

#
# exporting object - tabular
#

frequencies <- c(92,89,94,72,59) 

write.table(frequencies, 
            "frequencies.csv", 
            sep = ",", 
            row.names = FALSE, 
            col.names = TRUE, 
            fileEncoding = "UTF-8") 

#
# Corpus
#

require("tm")

my.dir <- "C:\\Users\\Lukas\\Desktop\\data\\LDA" 

directory.source <- DirSource(directory = my.dir, 
                              encoding = "UTF-8", 
                              ignore.case = T, 
                              pattern = ".txt") 

text.corpus <- Corpus(directory.source) 

#
# corpus operations
#

?removePunctuation()

?removeWords()

?stripWhitespace()

?removeNumbers()

?stemDocument()

?PlainTextDocument()

#
# corpus operations
#

edited.corpus <- text.corpus

edited.corpus <- tm_map(edited.corpus, removeNumbers)

edited.corpus <- tm_map(edited.corpus, removePunctuation)

edited.corpus <- tm_map(edited.corpus, stripWhitespace)

edited.corpus <- tm_map(edited.corpus, 
                        removeWords, 
                        stopwords("english"))

#
# Document term matrix
#

dtm <- DocumentTermMatrix(edited.corpus) 

dtm

dtm <- removeSparseTerms(dtm,sparse = 0.99)

dtm

dtm.matrixed <- as.matrix(dtm)

#
# LDA
#

library(topicmodels)

n.topics <- 10

lda.parameters <- list(verbose = 1,
                       iter = 500,
                       thin = 300,
                       burnin = 1000,
                       alpha = 50/n.topics)

model <- LDA(x = dtm,
             k = n.topics,
             method = "Gibbs",
             control = lda.parameters) 
#
# Exploring results - terms
#

terms(model,10)

model.terms <- terms(model,10)

#
# Exploring results - posterior distributions
#

model.posterior <- posterior(model)

model.posterior$topics

topic.doc.matrix <- model.posterior$topics 

model.posterior$terms

topic.terms.matrix <- model.posterior$terms

#
# exporting results as tabular data
#

write.table(topic.doc.matrix, 
            "topic.doc.matrix.txt", 
            sep = ",", 
            row.names = TRUE, 
            col.names = TRUE, 
            fileEncoding = "UTF-8")