eda with regression

Event-Data-Analysis.Rmd

@@ -1024,3 +1024,229 @@ dataWranglingfn <- function(term_list) {
 dataWranglingfn(term_list)
 ```
 
+```{r}
+# Load A BUNCH of packages
+library(shiny)
+library(shinydashboard)
+library(shinydashboardPlus)
+library(dplyr)
+library(tidyverse)
+library(reactable)
+library(reactablefmtr)
+library(shinycssloaders)
+library(lubridate)
+library(ggplot2)
+library(shinyjs)
+library(viridis)
+library(viridisLite)
+library(purrr)
+library(stringr)
+library(forcats)
+library(gganimate)
+library(ggthemes)
+library(leaflet)
+library(patchwork)
+library(probably)
+library(readr)
+library(janitor)
+library(shinyalert)
+library(shinyWidgets)
+library(tidyr)
+library(tidymodels)
+library(yardstick)
+library(plotly)
+library(hardhat)
+library(vip)
+library(rpart.plot)
+library(ranger)
+library(broom)
+library(googlesheets4)
+library(DT)
+library(sodium)
+library(formattable)
+library(shinyfullscreen)
+remotes::install_github("timelyportfolio/dataui")
+gs4_auth(email = "[email protected]", cache = ".secrets")
+
+
+
+## Local version
+# Define the directory containing the files
+data_dir <- "data/"
+
+term_list <- read_csv("data/term_list.csv")
+
+guest_only_base <- read_csv("data/guest_only_base.csv")
+
+dataWranglingfn <- function(term_list) {
+  # Create a function to read a file given its code
+  read_and_clean_event_file <- function(code) {
+    file_path <- paste0(data_dir, "2015-2024 Events Data - ", code, " - Event Data.csv")
+    data <- read_csv(file_path)
+    clean_names(data) %>%
+      mutate(across(everything(), as.character)) %>%
+      mutate(term = code)
+  }
+
+  term_to_year <- function(term) {
+    year <- as.numeric(str_sub(term, 2, 3))
+    season <- str_sub(term, 1, 1)
+    start_year <- if_else(season == "F", 2000 + year, 2000 + year - 1)
+    end_year <- start_year + 1
+    return(paste0(start_year, "-", end_year))
+  }
+
+  # Use purrr to read all files and store them in a named list
+  event_data_list <- set_names(map(term_list$term, read_and_clean_event_file), term_list$term)
+
+  combined_data <- reduce(event_data_list, full_join)
+
+  # Function to calculate the week of term
+  calculate_week_of_term <- function(event_date, term) {
+    start_date <- term_list %>%
+      filter(term == !!term) %>%
+      pull(start_date)
+
+    if(length(start_date) == 0) return(NA_integer_)  # Return NA if no start_date is found
+
+    # Calculate the week of the term based on Monday as the first day of the week
+    week_of_term <- as.integer((floor_date(event_date, unit = "week", week_start = 1) -
+                                  floor_date(start_date, unit = "week", week_start = 1)) / 7) + 1
+    return(week_of_term)
+  }
+
+  # Apply the new function to calculate week_of_term dynamically
+
+  combined_data_filtered <- combined_data %>%
+    filter(!is.na(what), what != "") %>%
+    filter(what != "Choir & Jazz Rehearsal") %>%
+    filter(what != "Jazz Rehearsal") %>%
+    mutate(date = as.Date(ymd(date))) %>%
+    mutate(
+      livestream = coalesce(livestream, live_stream)  # If 'livestream' is NA, use 'live_stream'
+    ) %>%
+    select(-live_stream) %>%
+    mutate(department = ifelse(str_detect(what, "CSA|Just Cellin|Lunar New Year|ACA|A Cappella|Accidentals|Exit 69|Date Knight|Knights|Knightingales|International Festival"), "CSA", department)) %>%
+    mutate(
+      support_level = fct_relevel(factor(support_level), "H", "M", "L"),
+      audio_needs = fct_relevel(factor(audio_needs), "H", "M", "L"),
+      stage_needs = fct_relevel(factor(stage_needs), "H", "M", "L"),
+      lighting_needs = fct_relevel(factor(lighting_needs), "H", "M", "L"),
+      projection = fct_relevel(factor(projection), "Y", "N"),
+      video_recording = fct_relevel(factor(video_recording), "Y", "N"),
+      livestream = fct_relevel(factor(livestream), "Y", "N"),
+      poster = fct_relevel(factor(poster), "Y", "N"),
+      program = fct_relevel(factor(program), "Y", "N"),
+      reception = fct_relevel(factor(reception), "Y", "N")
+    ) %>%
+    mutate(
+      venue = factor(venue),
+      department = factor(department)
+    ) %>%
+    mutate(
+      audience_count = as.numeric(ifelse(grepl("^[0-9]+$", audience_count), audience_count, NA)),
+      days_committed = as.numeric(ifelse(grepl("^[0-9]+$", days_committed), days_committed, NA)),
+      av_staff = as.numeric(ifelse(grepl("^[0-9]+$", av_staff), av_staff, NA)),
+      pac_staff = as.numeric(ifelse(grepl("^[0-9]+$", pac_staff), pac_staff, NA))
+    ) %>%
+    mutate(support_level = if_else(support_level == "N" | support_level == "Y", "L", support_level)) %>%
+    mutate(
+      department = str_replace_all(department, "WCC", "ODOA"),
+      department = str_replace_all(department, "MSUC", "MUSC"),
+      department = str_replace_all(department, "French Dept|French", "FREN"),
+      department = str_replace_all(department, "English", "ENGL"),
+      department = str_replace_all(department, "Pres. Office", "PRES"),
+      department = str_replace_all(department, "History", "HIST"),
+      department = str_replace_all(department, "THD", "THDA"),
+      department = str_replace_all(department, "Inclusion & Equity", "IEC"),
+      venue = str_replace_all(venue, "Skinner Chapel", "Chapel"),
+      department = str_replace_all(department, "/", " & "),
+      department = str_replace_all(department, ",", " &"),
+      venue = str_replace_all(venue, ",", " &"),
+      department = str_replace_all(department, "\\s+", " ")  # Remove extra spaces
+    ) %>%
+    select(-wk) %>%
+    mutate(department_type = case_when(
+      department == "MUSC" ~ "MUSC",
+      department == "ODOA" ~ "ODOA",
+      department == "CSA" ~ "CSA",
+      str_detect(department, "&") ~ "Collab",
+      TRUE ~ "Others"
+    )) %>%
+    mutate(what = str_replace_all(what, "Jazz Ensemble Concert|Jazz Area Concert", "Jazz Concert"),
+           what = str_replace_all(what, "Symphony Band Concert", "Symphony Concert"),
+           what = str_replace_all(what, "Composition Recital", "Composition Showcase Recital"),
+           what = str_replace_all(what, "Harpichord", "Harpsichord"),
+           what = str_replace_all(what, "Emsemble", "Ensemble"),
+           what = str_replace_all(what, "Juest Cellin'", "Just Cellin'"),
+           what = str_replace_all(what, "Facutly|FACULTY|Mazariello", "Faculty")) %>%
+    mutate(event_type = case_when(
+      str_detect(what, "GUEST|ODOA|Concert Series|SPCO") ~ "Guest / Masterclass",
+      str_detect(what, "Faculty") ~ "Faculty Recital",
+      str_detect(what, "Student|Senior|Junior|Piano Recital: |Johnson|Verma Jameson") ~ "Student Recital",
+      str_detect(what, "Studio Recital|Organ & Harpsichord|Composition Showcase Recital|Chamber Recital|Chamber Music Recital|Chamber Music|Organ Recital|Strings Recital|Violin & Viola|Violin/Viola|Drum Ensemble|Drum Recital|Voice Showcase Recital|Chinese Music Recital|Piano Studios Recital|Jazz Chamber|Piano Recital|Comps Fest|Recorder Recital|Music Ensemble|Studio") ~ "Studio Recital",
+      str_detect(what, "Orchestra Concert|Jazz Concert|Symphony Concert|Symphony Band|Choir Concert|Orchestra and Choir|Chinese & Global|Chinese Global Concert|Chinese and Global|Chinese Music Concert|Chinese Music Ensemble|Chinese Ensemble|Music Comps|Jazz Vocal Concert") ~ "Ensemble Concert",
+      str_detect(what, "CSA|Just Cellin|Lunar New Year|ACA|A Cappella|Accidentals|Exit 69|Date Knight|Knights|Knightingales|International Festival") ~ "Student Activity",
+      str_detect(what, "Masterclass|Lecture|Symposium") ~ "Guest / Masterclass",
+      str_detect(what, "Trustees|Trustee's|Presidents|Conference|President's|Presentation") ~ "Presentation",
+      str_detect(what, "Clinic|Music Fest|Music Department Showcase|Melinda Russell|Launch|Event|Opening") ~ "Special Events",
+      TRUE ~ "Guest / Masterclass"
+    )) %>%
+    mutate(event_type = case_when(str_detect(what, "Faculty&Guest") ~ "Faculty + Guest",
+                                  TRUE ~ event_type  # Keep existing values for other cases
+                                  )) %>% 
+    mutate(year = term_to_year(term)) %>%
+    mutate(term = factor(term, levels = term_list$term, ordered = TRUE)) %>%
+    arrange(year, term) %>%
+    mutate(
+      term_category = case_when(
+        str_detect(term, "^F") ~ "Fall",
+        str_detect(term, "^W") ~ "Winter",
+        str_detect(term, "^S") ~ "Spring"
+      )
+    ) %>%
+    mutate(term_category = factor(term_category, levels = c("Spring", "Winter", "Fall"), ordered = TRUE)) %>%
+    # Apply week_of_term calculation
+    rowwise() %>%
+    mutate(week_of_term = calculate_week_of_term(date, term)) %>%
+    ungroup()
+
+  # Example event summary after filtering and transformation
+  event_summary <- combined_data_filtered %>%
+    group_by(year, term) %>%
+    summarize(term_total = n(), .groups = 'drop') %>%
+    group_by(year) %>%
+    mutate(year_total = sum(term_total)) %>%
+    ungroup()
+
+  # combined_data_filtered <- combined_data_filtered %>%
+  #   left_join(guest_only_base %>% select(venue, date, what, department, genre, sponsor), by = join_by(venue, date, what, department)
+  #             # %>% select(what, genre, sponsor), by = "what", relationship = "many-to-many"
+  #             )
+  # Perform the join
+  combined_data_filtered_updated <- combined_data_filtered %>%
+    left_join(guest_only_base %>% select(venue, date, what, department, term, genre, sponsor), 
+              by = c("venue", "date", "what", "department", "term"))
+  
+
+  combined_data_filtered_updated %>% filter(event_type == "Guest / Masterclass",
+                                    str_detect(department, "MUSC")) %>%
+    filter(!str_detect(what, "Masterclass")) -> guest_only
+
+  # Get the unique years and append the "All" option
+  year_choices <- c("All", combined_data_filtered_updated %>% pull(year) %>% unique())
+
+  return(list(combined_data_filtered = combined_data_filtered_updated,
+              event_summary = event_summary,
+              year_choices = year_choices,
+              guest_only = guest_only))
+}
+
+dataWranglingfn(term_list) -> test1
+
+test1$combined_data_filtered %>% drop_na(audience_count, av_staff, pac_staff) -> combined_data_filtered_cleaned
+
+concert_lm <- lm(audience_count ~ support_level + av_staff*pac_staff, data = combined_data_filtered_cleaned)
+summary(concert_lm)
+```
+