In stormwaterheatmap/s8Regressions: What the Package Does (One Line, Title Case)

knitr::opts_chunk$set(echo=FALSE, 
  comment = '', fig.width = 6, fig.height = 6
)



library(tidyverse)
library(pastecs)
library(factoextra)
library(cluster)
library(DataExplorer)
library(tidymodels)
library(caret)
library(broom)
library(knitr)
library(kableExtra)
library(gtsummary)
load("~/GitHub/s8Regressions/data/s8_data.rda")


ps <- s8_data$paramgroup %>% unique()

s8_data$paramgroup %>% unique()

library(rlist)
s8_data<- s8_data %>% as.data.frame()
study_names <- s8_data$study_name
study_names_list <- s8_data$study_name %>% unique() %>% sort()

study_id <- study_names_list %>%  as.factor() %>% as.numeric()
types <- s8_data$type %>% unique()

study_names_short <- c('sea','tac','clark','king','pierce','pos','pot','sno')


lookup <- setNames(as.list(study_names_short), study_names_list)

s8_data = s8_data %>% 
  mutate(short_name =dplyr::recode(study_names, !!!lookup)) %>% 
  mutate(site_name = paste(sep = "-",dplyr::recode(study_names, !!!lookup),type)) %>% 
   filter(short_name != "clark") %>% 
  filter(sample_matrix %in% c("water", "Water", "WATER")) %>% 
   filter(paramgroup %in% c("Metal"))


s8_data %>% select(c("site_name","parameter","new_result_value")) %>% 
  pivot_wider(names_from = site_name, values_from= new_result_value, values_fn = length) %>%  kable() %>% kable_classic(full_width = F)

short_names = unique(s8_data$short_name)
site_names =  unique(s8_data$site_name)
i = 1

clus <- function(i) {


data.df <- s8_data %>% 
  #Metal", "Nutrient","Conventional")) %>%  #Conventional","Nutrient")) %>% #Nutrient","Conventional")) %>%
#filter(short_name ==short_names[i]) %>% 
  filter(site_name ==site_names[i]) %>% 
  filter(!parameter %in% c(
    "Biochemical Oxygen Demand - Water - Total", 
    #"Magnesium - Water - Dissolved", "Chloride - Water - Total",
    "Total Suspended Solids - Water - Total","Conductivity - Water - Total","Turbidity - Water - Total",
    "Ammonia - Water - Total",
    "Mercury - Water - Total", "Mercury - Water - Dissolved", #"Arsenic - Water - Dissolved",
    "Calcium - Water - Dissolved", "pH - Water - Total", "Calcium - Water - Total", "Hardness as CaCO3 - Water - Total" #"Magnesium - Water - Total"
  )) %>%
  filter(sample_matrix %in% c("water", "Water", "WATER")) %>%
  mutate(loc_date = paste0(site_name, "_",field_collection_start_date)) %>%
  pivot_wider(
    names_from = loc_date, id_cols = parameter,
    values_from = new_result_value,
    values_fn = mean
  ) %>%
  remove_rownames() %>%
  column_to_rownames("parameter") %>%
  drop_na() %>%
  t() %>%
  as.data.frame() %>%
  log() %>% 
  scale(center = TRUE, scale = TRUE) %>% 
  t() %>%
  as.data.frame() 


data.df2 <- data.df %>%

  mutate_all(function(x) as.numeric(as.character(x))) 




# 
km <- kmeans(data.df2, 3)
# km <- fanny(data.df2)

#return(fviz_nbclust(data.df2, kmeans, method = "silhouette"))

 return(
fviz_cluster(km,main = site_names[i],subtitle = paste("n =", ncol(data.df2)), data = data.df2,repel =TRUE,show.clust.cent = FALSE,ellipse.type = "confidence", ellipse.level = 0.90)
)
# # outplots[i] <- fviz_cluster(km.res1, data = data.df2,repel =TRUE,ellipse.type = "convex",show.clust.cent = FALSE)

}

#clus(1)
clus(2)
clus(3)
clus(4)
clus(5)
#clus(6)
clus(7)
#clus(8)
clus(9)
clus(10)
clus(11)
#```
#```r
clus(12)
clus(13)
clus(14)
#clus(15)
#clus(16)
#clus(17)
#clus(18)
#clus(19)