history/data_match.R

In fdrennan/test: What the Package Does (One Line, Title Case)

library(openxlsx)
library(tidyverse)
library(emmeans)
library(lme4)
library(nlme)
library(plotly)
library(ggpubr)
library(rlist)

files = list.files('R')
for (file in files) {
  source(paste0('R/', file))
}
pre_modeling <- function(input_data) {
  analysis_data <- data_clean(input_data)
  ready_final_model <- transform_diagnostics(analysis_data)
  return(ready_final_model)
}

# quickplot <- function(ready_final_model)
#   transformed_data  = ready_final_model$transformed_data
# if (all(is.na(transformed_data$Baseline))) {
#   var <- 'Response_Transformed'
#   plots <- vizualization(transformed_data = transformed_data,
#                          power =  power)
# } else{
#   var <- 'Response_Transformed_bc'
#   transformed_data <- transformed_data %>%
#     mutate(Response_Transformed_bc = Response_Transformed - Baseline)
#   debug(vizualization_cb)
#   plots <- vizualization_cb(transformed_data = transformed_data,
#                             power =  power)
# }


# final_modeling <- function(input_data, ready_final_model) {
#   var = ready_final_model$var
#   power = ready_final_model$box_cox
#   transformed_data = ready_final_model$transformed_data
#   transformed_data <-
#     variance_test_basic(transformed_data = transformed_data,
#                         variable = var)
#   transformed_data <-
#     variance_check(transformed_data, variable = var)
#   best_model <- cor_select(transformed_data, variable = var)
#   time_order = unique(input_data$Time)
#   final_model <- final_model(transformed_data = transformed_data,
#                              best = best_model, variable = var)
#   toi = 'Week6' #Need to add a column in the app to save timeSelection
#   contrast_list = generate_contrasts(toi = toi,
#                                      data = transformed_data,
#                                      time_order = time_order)
#   contrasts_stats = contrast_padjust(model = final_model,
#                                      contrast_list = contrast_list,
#                                      data = transformed_data)
#   #debug(final_output)
#   print(power)
#   output_tables = final_output(
#     transformed_data = transformed_data,
#     toi = toi,
#     emmeans_obj = contrasts_stats$emmeans_obj,
#     final_contrast = contrasts_stats$final_contrast,
#     power = power,
#     variable = var
#   )
#   return(output_tables)
# }

#input_data = read.csv('analysis_input_full.csv')[,-1]

input_data = read.csv('analysis_input_full_baseline.csv')
step_1 = pre_modeling(input_data)
quickplot(step_1) #Modify this function for an interactive shiny
last_step = final_modeling(input_data, step_1)