## Import data from CSV file
suppressMessages(library(tidyverse)) # required library
data <- read_csv("/kaggle/input/the-data-setlist/survey_data.csv", show_col_types = FALSE)

# Required library
library(dplyr)

## Create functions to be used

# Create horizontal bar plot using base R graphics    
suppressMessages(library(rlang, verbose=FALSE, quiet=TRUE)) # required library               
horiz_barplot <- function(data,
                          variable,
                          summ_var = NULL,
                          summ_fun = "n",
                          na.rm = TRUE,
                          others = FALSE,
                          others.p = NULL,
                          mar = c(5.1, 4.1, 4.1, 2.1),
                          col = "steelblue",
                          title = NULL,
                          xlab = NULL,
                          ylab = NULL,
                          xlim_adj = 0,
                          ...) {

  # --- Data Summarization ---

  # Use dplyr::count() for simple counts.
  if (is.null(summ_fun) || summ_fun == "n" || is.null(summ_var)) {
    summary_col_name <- "Count"
    
    summarized_data <- data %>%
      count(.data[[variable]], name = summary_col_name, sort = TRUE)

  } else {
    # Handle other summary functions (mean, median, etc.)
    summary_col_name <- paste(summ_fun, summ_var, sep = "_")
    
    summarized_data <- data %>%
      group_by(.data[[variable]]) %>%
      summarize(
        !!summary_col_name := get(summ_fun)(.data[[summ_var]], na.rm = na.rm),
        .groups = "drop"
      ) %>%
      arrange(desc(.data[[summary_col_name]]))
  }
  
  # --- "Others" Grouping ---
  if (others && !is.null(others.p)) {
    
    # Identify which rows to keep and which to group into "Others"
    to_keep <- summarized_data %>% filter(.data[[summary_col_name]] > others.p)
    to_group <- summarized_data %>% filter(.data[[summary_col_name]] <= others.p)
    
    # Only proceed if there are categories to group
    if (nrow(to_group) > 0) {
      others_row <- tibble(
        # Set the category name to "Others"
        !!variable := "Others",
        # Sum the values of the grouped categories
        !!summary_col_name := sum(to_group[[summary_col_name]])
      )
      
      # Combine the rows to keep with the new "Others" row
      summarized_data <- bind_rows(to_keep, others_row)
    }
  }

  # --- Prepare Data for Base R Plotting ---
  # `barplot` requires a named vector.
  
  # Ensure the variable column is character for naming
  summarized_data[[variable]] <- as.character(summarized_data[[variable]])
  
  # Create the named vector
  summary_vec <- setNames(summarized_data[[summary_col_name]], summarized_data[[variable]])
  
  # Reverse the vector so the largest bar is at the top of the plot
  summary_vec <- rev(summary_vec)

  # --- Generate Base R Plot ---
  
  # Set graphical margins
  par(mar = mar)
  
  # Display the bar plot
  bar_midpoints <- barplot(
    height = summary_vec,
    col = col,
    horiz = TRUE,
    las = 1, # Make category labels horizontal
    xlab = xlab,
    ylab = ylab,
    xlim = c(0, max(summary_vec) + xlim_adj),
    ...
  )
  
  # Add title and text labels for the bar values
  title(main = title, adj = 0, ...)
  text(
    x = summary_vec,
    y = bar_midpoints,
    labels = round(summary_vec, 1), # Round labels for cleanliness
    pos = 4, # Position labels to the right of the bars
    ...
  )
  
  # Return the bar midpoints, which can be useful
  invisible(bar_midpoints)
}

# Create custom mode function
mode <- function(x) {
  uniq_vals <- unique(x)
  uniq_vals[which.max(tabulate(match(x, uniq_vals)))]
}
              
            
## Data Understanding

# View the data
head(data)
tail(data)

# View column names
colnames(data)

# Check for missing values
paste("𝗡𝘂𝗺𝗯𝗲𝗿 𝗼𝗳 𝗡𝗔 𝘃𝗮𝗹𝘂𝗲𝘀:", sum(is.na(data)))


## Data Preparation

# Define the standardization function for program variable
standardize_program <- function(input_text) {
  cleaned <- str_to_title(input_text)
    
  cleaned <- str_replace_all(cleaned, "Bs ", "BS ")
  cleaned <- str_replace_all(cleaned, "Vs ", "BS ")
  cleaned <- str_replace_all(cleaned, "Ba ", "BA ")
  cleaned <- str_replace_all(cleaned, " Of ", " of ")
  cleaned <- str_replace_all(cleaned, " In ", " in ")
  cleaned <- str_replace_all(cleaned, " And ", " and ")
  cleaned <- str_replace_all(cleaned, "BS in", "BS")
  cleaned <- str_replace_all(cleaned, "Bachelor of Science in", "BS")
  cleaned <- str_replace_all(cleaned, "Bachelor of", "B")

  if (str_detect(cleaned, "^BA Social Science") &&
      str_detect(cleaned, "Economics")) {
    return("BA Social Sciences (Economics)")
  } else if (str_detect(cleaned, "^B Secondary Education")) {
    return("B Secondary Education")
  } else if (str_detect(cleaned, "^BS Biology")) {
    return("BS Biology")
  } else if (str_detect(cleaned, "^Bfa")) {
    return("B Fine Arts")
  } else {
    return(cleaned)
  }
}

# Clean the data
data <- data %>%
    mutate(
        program = sapply(program, standardize_program), # Standardize program column
        constituent = str_replace(constituent, "Banos", "Baños"), # "Banos" to "Baños" in constituent column
        year_standing = factor(year_standing, levels = c("I", "II", "III", "IV", "V"), ordered = TRUE)  
    )        


## Feature Understanding

# Create a dataframe of descriptions for student information
studentInfo_descr <- data.frame(
    variable = c("id", "constituent", "program", "year_standing", "lms"),
    description = c(
        "Unique identification number assigned to a student participant",
        "UP constituent university where a student was enrolled at the time of the survey",
        "Degree program of a student at the time of the survey",
        "Year level standing of a student at the time of the survey",
        "Learning Management System/s used by a student in virtual classes at the time of the survey"
    )
)

# Subset data of student information
studentInfo_wide <- data[, studentInfo_descr$variable]
              
# Create a dataframe of descriptions for UPSSS variables
upsss_descr <- data.frame(
    item_code = c(
        "upsss1", "upsss2", "upsss3", "upsss4", "upsss5", "upsss6", "upsss7", "upsss8", 
        "upsss9", "upsss10", "upsss11", "upsss12", "upsss13", "upsss14", "upsss15", 
        "upsss16", "upsss17", "upsss18", "upsss19", "upsss20"
    ),

    variable = c(
      "lms_org", "lms_stable", "lms_prog_track",
      "prof_content", "prof_tools_eff",
      "course_workload", "course_announce",
      "course_req_time", "course_feedback_time",
      "sync_stable", "prof_compassion",
      "lib_content", "lib_update", "uni_services",
      "reg_smooth", "enlist_success", "sem_length",
      "uni_response", "setup_conducive", "sat_overall"
    ),
    
    short_description = c(
        "Organization of the LMS", "Stability and reliability of the LMS", "LMS capability for progress tracking",
        "Provision of necessary learning content", "Instructors' effective use of tools",
        "Reasonableness of the course workload", "Timeliness of course announcements",
        "Sufficient time to complete requirements", "Timeliness of feedback on submissions",
        "Stability of synchronous classes", "Instructors' compassion for student concerns",
        "Availability of needed online library content", "Timeliness of online library content", "Utility of university-provided services", 
        "Smoothness of the registration process", "Success in enlisting required courses", "Adequacy of the semester's length",
        "University's response to student concerns", "Conduciveness of the remote learning setup", "Overall satisfaction with remote learning setup"
    ),
    
    full_description = c(
        "Most of my professors organized a platform (e.g. UVLe, Google Classroom, Canvas, etc.) in such a way that I can easily find the content that I need.",
        "Most of my professors used a platform (e.g. UVLe, Google Classroom, Canvas, etc.) that is reliable, stable, and secure.",
        "Most of my professors used a platform (e.g. UVLe, Google Classroom, Canvas, etc.) for me to keep track of my learning progress.",
        "Most of my professors provided the content that I need to learn the lessons (e.g. lecture notes, zoom recordings, video lectures, etc.).",
        "In general, my professors utilized the available tools and platforms effectively.",
        "Most of my professors assigned a reasonable amount of course requirements.",
        "Most of my professors announced the details of requirements according to the date stated in the course schedule.",
        "Most of my professors gave a sufficient amount of time to accomplish the course requirements.",
        "Most of my professors provided feedback to my submissions in a justifiable time.",
        "Most of my professors carried out synchronous classes in a stable and smooth manner.",
        "Most of my professors were compassionate in hearing the students' concerns about their courses and current well-being.",
        "I accessed the university's online library which provides the content that I need (e.g. journals, references, etc.).",
        "I accessed the university's online library which provides up-to-date content.",
        "I used the available services provided by the university in my study (e.g. UP mail can be used to access free Office, Canva for Education, unlimited Google Drive storage, etc.).",
        "The university's registration process was smooth and straightforward.",
        "I enlisted all the subjects that I needed in this semester.",
        "The semester's length is adequate.",
        "The university hears the students' concerns when it comes to issues such as academic ease due to our current situation.",
        "I believe the university's remote learning setup was organized in a way that is conducive to learning.",
        "As a whole, I am satisfied with the university's remote learning setup this semester."
    )
)

# Subset UPSSS data  
upsss_wide <- data[, names(data) == "id" | grepl("upsss", names(data))]
upsss_items <- (colnames(data))[grepl("upsss", colnames(data))]
              
# Required libraries
suppressMessages(library(tidyr, verbose=FALSE, quiet=TRUE))
suppressMessages(library(reshape2, verbose=FALSE, quiet=TRUE))

# Reshape data
upsss_long <- melt(
        upsss_wide,
        id.vars="id",
        variable.name="item_code",
        value.name="rating"
    ) %>%
    left_join(upsss_descr, by="item_code") %>%
    mutate(item_code = factor(item_code, levels=upsss_items, ordered = TRUE)) %>%
    inner_join(studentInfo_wide, "id")

# Create a dataframe of descriptions for SPAPS variables
spaps_descr <- data.frame(
    item_code = c(
        "spaps1", "spaps2", "spaps3", "spaps4", "spaps5", "spaps6", "spaps7"
     ),

    variable = c(
        "req_submit_time",
        "critical_thinking",
        "feedback_exchange",
        "feedback_apply",
        "course_obj_achieve",
        "study_daily",
        "gwa_expected"
    ),
    
    short_description = c(
        "Timely submission of requirements",
        "Application of critical thinking",
        "Providing and receiving feedback", 
        "Application of received feedback", 
        "Achievement of course objectives",        
        "Maintenance of a daily study habit",
        "Expectation of high GWA (1.75 or better)"        
    ),
    
    full_description = c(
        "I was able to submit all my requirements this semester on time.",
        "I was able to think critically and logically while doing my requirements.",
        "I was able to provide and receive feedback on my requirements.",
        "I was able to apply the feedback that I received.",
        "I was able to achieve all of my courses’ objectives.",
        "I made sure that I study for my courses every day.",
        "I believe I can get a general weighted average (GWA) of at least 1.75 this semester."
    )
        
)

# Subset SPAPS data  
spaps_wide <- data[, names(data) == "id" | grepl("spaps", names(data))]
spaps_items <- (colnames(data))[grepl("spaps", colnames(data))]

# Reshape data
spaps_long <- melt(
        spaps_wide,
        id.vars="id",
        measure.vars=spaps_items,
        variable.name="item_code",
        value.name="rating"
    ) %>%
    left_join(spaps_descr, by="item_code") %>%
    mutate(item_code = factor(item_code, levels = spaps_items, ordered=TRUE)) %>%
    inner_join(studentInfo_wide, "id")

# Display bar plot
# Constituent university
horiz_barplot(
  data = data,
  variable = "constituent",
  title = "\nNumber of Students\nby Constituent University",
  xlab = "Number of Students",
  mar = c(5, 10, 2, 0),
  xlim_adj = 20
)

# Display bar plot
# Undergraduate program
horiz_barplot(
  data = data,
  variable = "program",
  others = TRUE,
  others.p = 1,
  title = "\nNumber of Students\nby Undergraduate Program",
  xlab = "Number of Students",
  mar = c(5, 22, 2, 0),
  xlim_adj = 20
)

# Display bar plot
# Undergraduate program
horiz_barplot(
  data = data,
  variable = "year_standing",
  others = TRUE,
  others.p = 1,
  title = "\nNumber of Students\nby Year-Level Standing",
  xlab = "Number of Students",
  mar = c(5, 2, 2, 0),
  xlim_adj = 20
)

# Aggregate student ratings
agg_data <- upsss_long %>%
    group_by(id) %>%
    summarize(upsss_ratingSum=sum(rating)) %>%
    inner_join(
        spaps_long %>%
            group_by(id) %>%
            summarize(spaps_ratingSum=sum(rating)),
        by="id"
    ) %>%
    inner_join(studentInfo_wide, by="id")

# Compute median
med_upsss_ratingSum <- median(agg_data$upsss_ratingSum)

# Display summary outputs
data.frame(med_ratingSum = med_upsss_ratingSum, avg_med_ratingSum = med_upsss_ratingSum/20)

# Display summary statistics
upsss_long %>%
    group_by(item_code, variable) %>%
    summarize(n_students = n(),
              #mean_rating = mean(rating),
              min_rating = min(rating), 
              med_rating = median(rating),
              mode_rating = mode(rating),
              max_rating = max(rating),         
              iqr = IQR(rating),
              .groups = "keep"  
    ) %>%
    arrange(desc(med_rating), variable)

# Display bar plot
horiz_barplot(
  data = upsss_long,
  variable = "variable",
  summ_var="rating",
  summ_fun="median",
  title = "\nMedian Rating\nby UPSSS Variable",
  xlab = "Median Rating",
  mar = c(5, 10, 2, 0),
  xlim_adj = 0.5
)

# Display summary statistics
upsss_long %>%
    group_by(constituent) %>%
    summarize(n_students = n()/20,
              min_upsss_rating = min(rating), 
              med_upsss_rating = median(rating),
              mode_upsss_rating = mode(rating),
              max_upsss_rating = max(rating),
              iqr = IQR(rating),
              .groups = "keep"  
    ) %>%
    arrange(desc(med_upsss_rating))

# Display bar plot
horiz_barplot(
  data = upsss_long,
  variable = "constituent",
  summ_var="rating",
  summ_fun="median",
  title = "\nMedian UPSSS Rating\nby Constituent University",
  xlab = "Median UPSSS Rating",
  mar = c(5, 10, 2, 0),
  xlim_adj = 0.5
)

# Display summary statistics
upsss_long %>%
    group_by(year_standing) %>%
    summarize(n_students = n()/20,
              min_upsss_rating = min(rating), 
              med_upsss_rating = median(rating),
              mode_upsss_rating = mode(rating),
              max_upsss_rating = max(rating),
              iqr = IQR(rating),
              .groups = "keep"  
    ) %>%
    arrange(desc(med_upsss_rating))

# Display bar plot
horiz_barplot(
  data = upsss_long,
  variable = "year_standing",
  summ_var="rating",
  summ_fun="median",
  title = "\nMedian UPSSS Rating\nby Year Level Standing",
  xlab = "Median UPSSS Rating",
  xlim_adj = 0.5
)

# Aggregate student ratings

# Compute median
med_spaps_ratingSum <- median(agg_data$spaps_ratingSum)

# Display summary outputs
data.frame(med_ratingSum = med_spaps_ratingSum, avg_med_ratingSum = round(med_spaps_ratingSum/7, 1))

# Display summary statistics
spaps_long %>%
    group_by(item_code, variable) %>%
    summarize(min_rating = min(rating), 
              med_rating = median(rating),
              max_rating = max(rating),
              iqr = IQR(rating),
              .groups = "keep"  
    ) %>%
    arrange(desc(med_rating), variable)

# Display bar plot
horiz_barplot(
  data = spaps_long,
  variable = "variable",
  summ_var="rating",
  summ_fun="median",
  title = "\nMedian Rating\nby SPAPS Variable",
  xlab = "Median Rating",
  mar = c(5, 10, 2, 0),
  xlim_adj = 2.5
)

# Display summary statistics
spaps_long %>%
    group_by(constituent) %>%
    summarize(n_students = n()/7,
              min_spaps_rating = min(rating), 
              med_spaps_rating = median(rating),
              mode_spaps_rating = mode(rating),
              max_spaps_rating = max(rating),
              iqr = IQR(rating),
              .groups = "keep"  
    ) %>%
    arrange(desc(med_spaps_rating))

# Display bar plot
horiz_barplot(
  data = spaps_long,
  variable = "constituent",
  summ_var="rating",
  summ_fun="median",
  title = "\nMedian SPAPS Rating\nby Constituent University",
  xlab = "Median SPAPS Rating",
  mar = c(5, 10, 2, 0),
  xlim_adj = 0.5
)

# Display summary statistics
spaps_long %>%
    group_by(year_standing) %>%
    summarize(n_students = n()/7,
              min_spaps_rating = min(rating), 
              med_spaps_rating = median(rating),
              mode_spaps_rating = mode(rating),
              max_spaps_rating = max(rating),
              iqr = IQR(rating),
              .groups = "keep"  
    ) %>%
    arrange(desc(med_spaps_rating))

# Display bar plot
horiz_barplot(
  data = spaps_long,
  variable = "year_standing",
  summ_var="rating",
  summ_fun="median",
  title = "\nMedian SPAPS Rating\nby Year Level Standing",
  xlab = "Median SPAPS Rating",
  xlim_adj = 0.5
)

## Test rank correlation of UPSSS and SPAPS rating sums

# Spearman
cor.test(
    x = agg_data$upsss_ratingSum,
    y = agg_data$spaps_ratingSum,
    method="spearman",
    exact=FALSE
)

# Kendall
cor.test(
    x = agg_data$upsss_ratingSum,
    y = agg_data$spaps_ratingSum,
    method="kendall"
)

## Display scatterplot
plot(
    x = agg_data$upsss_ratingSum,
    y = agg_data$spaps_ratingSum,
    pch=19,
    col="steelblue",
    main="Student Satisfaction vs. Perceived Academic Performance",
    xlab="UPSSS Rating Sum",
    ylab="SPAPS Rating Sum"
)

## (UPSSS vs. SPAPS) variables

# Join UPSSS and SPAPS data
upsss_spaps_wide <- upsss_wide %>%
    inner_join(spaps_wide, by="id") %>%
    select(-id)

# Rename columns
colnames(upsss_spaps_wide) <- c(upsss_descr$variable, spaps_descr$variable)

# Calculate the correlation matrices

# Spearman
spearmanCor_matrix <- upsss_spaps_wide %>%
  select(all_of(spaps_descr$variable), all_of(upsss_descr$variable)) %>%
  cor(method="spearman") 

# Kendall
kendallCor_matrix <- upsss_spaps_wide %>%
  select(all_of(spaps_descr$variable), all_of(upsss_descr$variable)) %>%
  cor(method="kendall")

# spearmanCor_matrix[c(8:27), c(1:7)]
# kendallCor_matrix[c(8:27), c(1:7)]

# Generate heatmaps

# Required library
# install.packages("pheatmap")
library(pheatmap)

# Spearman
pheatmap(
  spearmanCor_matrix[c(8:27), c(1:7)],
  main = "\nSpearman Rank Correlation Matrix\n(UPSSS vs. SPAPS)",
  #cluster_rows = FALSE,
  #cluster_cols = FALSE,
  display_numbers = TRUE
)

# # Kendall
# pheatmap(
#   kendallCor_matrix[c(8:27), c(1:7)],
#   main = "\nKendall Rank Correlation Matrix\n(UPSSS vs. SPAPS)",
#   #cluster_rows = FALSE,
#   #cluster_cols = FALSE,
#   display_numbers = TRUE
# )