2022_12_19.R

library(pacman)

`%notin%` <- Negate(`%in%`)

p_load(
  #--- Packages to Fit Models
  MASS, logistf, survival, ISLR2, tree, randomForest, gbm, nnet, caret,
  #--- Packages to Produce Tables
  gtsummary, flextable, janitor, broom, officer, kableExtra, reactable, 
  #--- Packages to Produce Figures
  crayon, ggsci, ggridges, ggthemes, ggforce, ggpubr, patchwork, grid, gridExtra, plotly,  survminer, viridis, ggridges, hrbrthemes, stickylabeller, latex2exp, scales, glue, 
  #--- Packages for Data Retrieval & Pre-Processing
  readxl, here, rdrop2, lubridate, zoo, tidyverse, purrr, data.table, stringr, tidyr, table1
)

data <- read_csv(file.path(here(), "data", "created_data", "2022_12_02.csv"))


## categorising proportion based on #of times/year. 
data <- data %>% 
  mutate(prop_cat = factor(case_when(prop <= 0.1 ~ "stage1", 
                                     (0.1 < prop & prop <= 0.30) ~ "stage2", 
                                     (0.30 < prop & prop <= 1.00) ~ "stage3"),
                           levels = c("stage1", "stage2", "stage3")), 
         sex = as.factor(sex), 
         n_br = as.factor(n_br), 
         n_dep = as.factor(n_dep), 
         name = as.factor(name))

# ## v basic tree fitting
# tree.naive.model <- tree(prop_cat ~ age + sex + n_dep + name, data = data)
# summary(tree.naive.model)
# 
# ## pruning the tree above
# cv.tree.naive.model <- cv.tree(tree.naive.model, FUN = prune.misclass)
# cv.tree.naive.model

## random forest 
# rf.naive.model <- randomForest(prop_cat ~ age + sex + n_dep + name, 
#                                data = data, 
#                                ntree = 30, 
#                                classwt = data %>% group_by(prop_cat) %>% summarise(n = n()) %>% pull(n), 
#                                importance = TRUE)
# rf.naive.model

## boosted trees
# data.boost <- data %>% mutate(prop_cat = case_when(prop_cat == "stage1" ~ 0, 
#                                                    prop_cat == "stage2" ~ 1))
# b.naive.model <- gbm(prop_cat ~ age + sex + n_dep + name, 
#                      data = data.boost, 
#                      distribution = "bernoulli", 
#                      n.trees = 5000, interaction.depth = 3)
# summary(b.naive.model)

confusionMatrix(predict(multinom(prop_cat ~ age + sex + n_dep + name, data = data),  data), 
                data$prop_cat)