library(tidyverse)
main_csv <-"~/analysis_data/120725_unified.csv"
main_df <- read.csv(main_csv, header = TRUE) 

dsl_csv <-"~/dsl/120725_DSL_frame.csv"
dsl_df <- read.csv(dsl_csv, header = TRUE) 

needs_triage <- dsl_df |>
  filter(week_index >= -4) |>
  filter(priority=="Needs Triage") |>
  mutate(ttr_weeks = TTR/168) |>
  group_by(source, isAuthorWMF)|>
  summarize(
    mean_ttr_weeks = mean(ttr_weeks),
    sd_ttr_weeks = sd(ttr_weeks)
  )
  
changes<- dsl_df |>
  filter(priority == "Needs Triage" | 
           priority == "Unbreak Now!" |
           priority == "High") |>
  mutate(
    period = case_when(
      week_index >= -4 & week_index <= 4   ~ "8 weeks after announcement",
      week_index >= -13 & week_index <= -5 ~ "8 weeks before deployment announcement",
      TRUE ~ NA
    )
  ) %>%
  filter(!is.na(period)) |>
  mutate(ttr_weeks = TTR/168) |>
  group_by(source, period, priority) %>%
  summarise(
    count = n(),
    mean_ttr_weeks = mean(ttr_weeks, na.rm = TRUE),
    sd_ttr_weeks = sd(ttr_weeks, na.rm = TRUE),
  )


# new contributors 
first_task <- main_df |>
  filter(comment_type == "task_description") |>
  group_by(source, AuthorPHID) |>
  summarise(
    task_count = n(),
    first_task_week = min(week_index)
  ) 

tasks_flagged <- main_df %>%
  filter(comment_type == "task_description") |>
  left_join(first_task, by = c("source", "AuthorPHID")) %>%
  mutate(is_first_time_author = week_index == first_task_week)

summary_df <- tasks_flagged %>%
  group_by(week_index, source) %>%
  summarize(
    total_tasks = n(),
    first_time_tasks = sum(is_first_time_author),
    proportion_first_time = first_time_tasks / total_tasks
  ) %>%
  ungroup()

period_counts <- summary_df %>%
  mutate(
    period = case_when(
      week_index >= -4 & week_index <= 4   ~ "recent",
      week_index >= -13 & week_index <= -5 ~ "prior",
      TRUE ~ NA_character_
    )
  ) %>%
  filter(!is.na(period)) %>%
  group_by(source, period) %>%
  summarize(period_first_time_tasks = sum(first_time_tasks), .groups = "drop")

ggplot(summary_df, aes(x = week_index, y = first_time_tasks)) +
  facet_grid(source ~ ., 
             scales = "free_y",
             labeller = labeller(source = c("c1" = "VisualEditor", 
                                            "c2" = "HTTPS-login", 
                                            "c3" = "HTTP-deprecation"))) + 
  geom_col() +
  labs(
    x = "Date of first task",
    y = "Number of tasks created",
    title = "Task count by Author's first task date"
  ) +
  theme_minimal()