animal-crossing.Rmd

---
title: "Animal Crossing"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

```{r}
library(tidyverse)
theme_set(theme_light())

critic <- readr::read_tsv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2020/2020-05-05/critic.tsv')
user_reviews <- readr::read_tsv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2020/2020-05-05/user_reviews.tsv')
items <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2020/2020-05-05/items.csv')
villagers <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2020/2020-05-05/villagers.csv')
```

```{r}
villagers
items %>%
  View()
```

### Text Analysis

```{r}
critic %>%
  ggplot(aes(grade)) +
  geom_histogram()

user_reviews %>%
  ggplot(aes(grade)) +
  geom_histogram()

head(user_reviews$text)
```

```{r}
library(tidytext)
library(lubridate)

user_review_words <- user_reviews %>%
  unnest_tokens(word, text) %>%
  anti_join(stop_words, by = "word") %>%
  count(user_name, date, grade, word)
```

```{r}
by_week <- user_reviews %>%
  group_by(week = floor_date(date, "week", week_start = 1)) %>%
  summarize(nb_reviews = n(),
            avg_grade = mean(grade),
            pct_zero = mean(grade == 0),
            pct_ten = mean(grade == 10))

by_week %>%
  ggplot(aes(week, avg_grade)) +
  geom_line() +
  geom_point(aes(size = nb_reviews)) +
  expand_limits(y = 0) +
  labs(x = "Time",
       y = "Average grade",
       size = "# of reviews")

user_reviews %>%
  ggplot(aes(grade)) +
  geom_histogram() +
  labs(title = "Most reviews were very low or very high")

by_week %>%
  gather(type, value, contains("pct")) %>%
  mutate(type = ifelse(type == "pct_zero", "% rated 0", "% rated 10")) %>%
  ggplot(aes(week, value, color = type)) +
  geom_line() +
  geom_point(aes(size = nb_reviews)) +
  scale_y_continuous(labels = scales::percent) +
  expand_limits(y = 0) +
  labs(x = "Time",
       y = "% of reviews",
       size = "Total reviews in week",
       title = "Reviews got more polarizing in middle of game")
```


```{r}
# user_reviews %>%
#   mutate(start_text = str_sub(text, 1, pmin(30, str_length(text))),
#          regex = paste0(".", start_text, ".*")) %>%
#   mutate(new_text = map2_chr(text, regex, str_remove)) %>%
#   filter(str_length(new_text) != str_length(text))

by_word <- user_review_words %>%
  group_by(word) %>%
  summarize(avg_grade = mean(grade),
            nb_reviews = n()) %>%
  arrange(desc(nb_reviews)) %>%
  filter(nb_reviews >= 25) %>%
  arrange(desc(avg_grade))
  
by_word %>%
  filter(nb_reviews >= 75) %>%
  ggplot(aes(nb_reviews, avg_grade)) +
  geom_point() +
  geom_text(aes(label = word), vjust = 1, hjust = 1, check_overlap = TRUE) +
  scale_x_log10()
```

```{r}
by_word %>%
  top_n(20, -avg_grade) %>%
  ggplot(aes(nb_reviews, avg_grade)) +
  geom_point() +
  geom_text(aes(label = word), vjust = 1, hjust = 1, check_overlap = TRUE) +
  scale_x_log10() +
  labs(title = "What words were associated with low-grade reviews?",
       subtitle = "20 most negative words; only words in at least 25 reviews")
```

```{r}
library(widyr)
library(stm)

review_matrix <- user_review_words %>%
  group_by(word) %>%
  filter(n() >= 25) %>%
  cast_sparse(user_name, word, n)

topic_model_6 <- stm(review_matrix,
                     K = 6,
                     verbose = TRUE,
                     init.type = "Spectral",
                     emtol = 5e-5)
```

```{r}
tidy(topic_model_6) %>%
  group_by(topic) %>%
  top_n(12, beta) %>%
  mutate(term = reorder_within(term, beta, topic)) %>%
  ggplot(aes(beta, term, fill = factor(topic))) +
  geom_col(show.legend = FALSE) +
  scale_y_reordered() +
  facet_wrap(~ topic, scales = "free_y")

tidy(topic_model_6) %>%
  filter(term == "progress")

topic_model_gamma <- tidy(topic_model_6, matrix = "gamma") %>%
  mutate(user_name = rownames(review_matrix)[document]) %>%
  inner_join(user_reviews, by = "user_name")

topic_model_gamma %>%
  group_by(topic) %>%
  top_n(1, gamma)

topic_model_gamma %>%
  group_by(topic) %>%
  summarize(correlation = cor(gamma, grade),
            spearman_correlation = cor(gamma, grade, method = "spearman"))

topic_model_gamma %>%
  group_by(week = floor_date(date, "week", week_start = 1),
           topic) %>%
  summarize(avg_gamma = mean(gamma)) %>%
  ggplot(aes(week, avg_gamma, color = factor(topic))) +
  geom_line() +
  expand_limits(y = 0) +
  scale_y_continuous(labels = scales::percent) +
  labs(x = "Time",
       y = "Average gamma (document-topic association)")
```