analysis.R

library(mongolite)

# Store your connection string in an .env or R environment variable
connection_string <- Sys.getenv("MONGODB_URI")

airbnb_collection <- mongo(
  collection = "listingsAndReviews",
  db = "sample_airbnb",
  url = connection_string
)

# Test connection
test <- airbnb_collection$find(limit = 5)

print(test)

#Now I wanna pull a focused dataset 
library(mongolite)
library(dplyr)
library(stringr)

airbnb_df <- airbnb_collection$find(
  query = '{}',
  fields = '{
    "name": 1,
    "summary": 1,
    "description": 1,
    "space": 1,
    "neighborhood_overview": 1,
    "notes": 1,
    "transit": 1,
    "access": 1,
    "interaction": 1,
    "house_rules": 1,
    "property_type": 1,
    "room_type": 1,
    "bed_type": 1,
    "minimum_nights": 1,
    "maximum_nights": 1,
    "accommodates": 1,
    "bedrooms": 1,
    "beds": 1,
    "bathrooms": 1,
    "number_of_reviews": 1,
    "price": 1,
    "cleaning_fee": 1,
    "extra_people": 1,
    "address.market": 1,
    "address.country": 1,
    "review_scores.review_scores_rating": 1,
    "_id": 0
  }'
)

glimpse(airbnb_df)

#To clean a bit 
# First i´m inspecting the nested columns
names(airbnb_df$address)
names(airbnb_df$review_scores)

head(airbnb_df$address)
head(airbnb_df$review_scores)
#Clean Dataframe
library(dplyr)

airbnb_clean <- airbnb_df %>%
  mutate(
    minimum_nights = as.numeric(minimum_nights),
    maximum_nights = as.numeric(maximum_nights),
    market = address$market,
    country = address$country,
    review_score = review_scores$review_scores_rating
  ) %>%
  select(
    name, summary, space, description, neighborhood_overview, notes,
    transit, access, interaction, house_rules,
    property_type, room_type, bed_type,
    minimum_nights, maximum_nights,
    accommodates, bedrooms, beds, bathrooms,
    number_of_reviews, price, cleaning_fee, extra_people,
    market, country, review_score
  )
glimpse(airbnb_clean)
colSums(is.na(airbnb_clean))

#Viz 1 -Price Distribution
library(ggplot2)
library(scales)

ggplot(airbnb_clean %>% filter(price < 1000), aes(x = price)) +
  geom_histogram(bins = 40) +
  scale_x_continuous(labels = dollar_format()) +
  labs(
    title = "Distribution of Airbnb Listing Prices (Filtered)",
    x = "Price",
    y = "Number of Listings"
  ) +
  theme_minimal()
#Viz 2- Price by Room type 
#used median instead of mean because airbnb prices are skewed, the mean is distored by luxury listings 
avg_price_room <- airbnb_clean %>%
  group_by(room_type) %>%
  summarise(median_price = median(price, na.rm = TRUE))

ggplot(avg_price_room, aes(x = reorder(room_type, median_price), y = median_price)) +
  geom_col() +
  coord_flip() +
  scale_y_continuous(labels = dollar_format()) +
  labs(
    title = "Median Price by Room Type",
    x = "Room Type",
    y = "Median Price"
  ) +
  theme_minimal()

#Viz 3 -Top Markets
top_markets <- airbnb_clean %>%
  count(market, sort = TRUE) %>%
  slice_head(n = 10)

ggplot(top_markets, aes(x = reorder(market, n), y = n)) +
  geom_col() +
  coord_flip() +
  labs(
    title = "Top Markets by Number of Listings",
    x = "Market",
    y = "Listings"
  ) +
  theme_minimal()

#Viz 4- Bedroom vs price 
avg_price_bedrooms <- airbnb_clean %>%
  filter(price < 1000, bedrooms <= 6) %>%
  group_by(bedrooms) %>%
  summarise(avg_price = mean(price, na.rm = TRUE))

ggplot(avg_price_bedrooms, aes(x = bedrooms, y = avg_price)) +
  geom_line() +
  geom_point() +
  scale_y_continuous(labels = dollar_format()) +
  labs(
    title = "Average Price by Bedrooms (Filtered)",
    x = "Bedrooms",
    y = "Average Price"
  ) +
  theme_minimal()

#double checking its clean
airbnb_text <- airbnb_clean %>%
  mutate(
    full_text = paste(
      name, summary, space, description,
      neighborhood_overview, notes,
      transit, access, interaction, house_rules,
      sep = " "
    )
  )

#tokenize 
library(tidytext)
library(tidyr)
library(stringr)

tidy_words <- airbnb_text %>%
  select(room_type, market, full_text) %>%
  unnest_tokens(word, full_text)
#cleaning words (removes the, and, is..)
data("stop_words")

tidy_words_clean <- tidy_words %>%
  anti_join(stop_words, by = "word") %>%
  filter(str_detect(word, "[a-z]")) %>%
  filter(nchar(word) > 2)

#WORD FREQUENCY 1
tidy_words_clean <- tidy_words %>%
  anti_join(stop_words, by = "word") %>%
  filter(str_detect(word, "[a-z]")) %>%
  filter(nchar(word) > 2) %>%
  filter(!word %in% c("apartment", "room", "bed", "bedroom", "house"))
top_words <- tidy_words_clean %>%
  count(word, sort = TRUE)

top_words %>%
  slice_head(n = 15) %>%
  ggplot(aes(x = reorder(word, n), y = n)) +
  geom_col() +
  coord_flip() +
  labs(
    title = "Most Frequent Words in Airbnb Listings",
    x = "Word",
    y = "Frequency"
  ) +
  theme_minimal()


#SENTIMENT ANALYSIS 2
bing_sentiment <- tidy_words_clean %>%
  inner_join(get_sentiments("bing"), by = "word")

sentiment_summary <- bing_sentiment %>%
  count(sentiment)

ggplot(sentiment_summary, aes(x = sentiment, y = n, fill = sentiment)) +
  geom_col() +
  labs(
    title = "Sentiment in Airbnb Listings",
    x = "Sentiment",
    y = "Word Count"
  ) +
  theme_minimal()
#sentiment by roomtype 
sentiment_by_room <- bing_sentiment %>%
  count(room_type, sentiment)

ggplot(sentiment_by_room, aes(x = sentiment, y = n, fill = sentiment)) +
  geom_col() +
  facet_wrap(~room_type) +
  labs(
    title = "Sentiment by Room Type"
  ) +
  theme_minimal()

#TF-IDF 3
tfidf_words %>%
  arrange(desc(tf_idf)) %>%
  group_by(room_type) %>%
  slice_head(n = 10) %>%
  ggplot(aes(x = reorder(word, tf_idf), y = tf_idf, fill = room_type)) +
  geom_col(show.legend = FALSE) +
  facet_wrap(~room_type, scales = "free") +
  coord_flip() +
  labs(
    title = "Top Terms by Room Type (TF-IDF)",
    x = "Word",
    y = "Importance"
  ) +
  theme_minimal() +
  theme(strip.text = element_text(size = 10))