Tidy Text Example in R Stripping out numeric, punc, and replace words

tidytext.R

library(dplyr)
library(tidytext)
library(ggplot2)

data <- read.csv('/Users/steven/Desktop/JEOPARDY_CSV.csv')
text_set <- iconv(data$Question, to = "utf-8-mac")

clean_text <- text_set %>%
  as_tibble() %>%
  unnest_tokens(word, value,  to_lower = TRUE, strip_punct = TRUE, strip_numeric = TRUE) %>%
  mutate(word = str_replace(word, "word-to-replace", "replacement-word")) %>%
  anti_join(stop_words) 
  
  # For multiple replacement words use this 
  # mutate(word = if_else(word %in% c("emailing", "emails"), "email", word))
  
  # For single replacement use this
  # mutate(word = str_replace(word, "word-to-replace", "replacement-word")) %>%
  
# Bing Word Counts to print out do print(bing_word_counts)
bing_word_counts <- clean_text %>%
  inner_join(get_sentiments("bing")) %>%
  count(word, sentiment, sort = TRUE) %>%
  ungroup()

# Use bing word counts to make a graph
bing_word_counts %>%
  group_by(sentiment) %>%
  top_n(10) %>%
  ungroup() %>%
  mutate(word = reorder(word, n)) %>%
  ggplot(aes(word, n, fill = sentiment)) +
  geom_col(show.legend = FALSE) +
  facet_wrap(~sentiment, scales = "free_y") +
  labs(y = "Contribution to sentiment",
       x = NULL) +
  coord_flip()


#  Notes 
# # custom stop words, to be removed from analysis
# custom_stop_words <- tibble(word = c("trump","critical","issues","issue"))
# 
# # To remove a list of extra stop_words add this to the first method
#   anti_join(custom_stop_words, by = "word")