R/importNCRNWater.R

In NCRN/NCRNWater: Data analysis and visualization for NPS I&M water monitoring

#' @include NCRNWater_Park_Class_def.R
#' @include NCRNWater_Site_Class_def.R
#' @include NCRNWater_Characteristic_Class_def.R
#' @title importNCRNWater
#' 
#' @description This function imports data from a .csv files exported from NPStoret and saves it as \code{Park} objects. 
#' 
#' @param Dir The directory where the data is found
#' @param Data The data file. Defaults to "Water Data.csv"
#' @param MetaData The metadata file. Defaults to "MetaData.csv"
#' @param wqx Indicates if the data file is in wqx format, defaults to \code{FALSE}.
#' 
#' @return Returns \code{Park} objects, one for each park, as a \code{list}.
#' 
#' @importFrom dplyr distinct group_by filter mutate rename select ungroup
#' @importFrom lubridate mdy ymd
#' @importFrom magrittr %>%
#' @importFrom purrr map map2 pmap
#' @importFrom readr cols read_csv
#' @importFrom methods new
#' 
#' @examples 
#' ncrnwd<-importNCRNWater(Dir = "./Data/NCRN", Data = "Water Data.csv", MetaData = "VizMetaData.csv")
#' 
#' @export


importNCRNWater<-function(Dir, Data="Water Data.csv", MetaData="MetaData.csv", wqx=FALSE){
  
#### Read in Data ####
  if(wqx){
    Indata <- read_csv(paste(Dir, Data, sep="/"), col_types=cols(.default="c")) %>% 
      rename(SiteCode = MonitoringLocationIdentifier, Date =`ActivityStartDate`, 
             Characteristic = `CharacteristicName`, Value = `ResultMeasureValue`)%>% 
      mutate(TextValue=Value)
    #### Get data ready to make into objects ####
    Indata$Date<-ymd(Indata$Date)
  }
  
  
  
 if(!wqx) {
  Indata <- read_csv(paste(Dir, Data, sep="/"), col_types=cols(.default="c")) %>% 
    rename(SiteCode = StationID, Date =`Visit Start Date`, Characteristic = `Local Characteristic Name`,
           Value = `Result Value/Text`)%>% 
    mutate(TextValue=Value)
  #### Get data ready to make into objects ####
  Indata$Date<-mdy(Indata$Date)
  
 }
  
  if(any(names(Indata)=="ValueCen") & any(names(Indata)=="Censored")){
    
    Indata <- Indata %>%  
      mutate(ValueCen = as.numeric(ValueCen), Censored = as.logical(Censored))
  }
    
  MetaData<-read_csv(paste(Dir, MetaData, sep="/"), col_types=cols()) #makes function less chatty
  


#### Check whether MQL and UQL fields (Minimum and Upper Detection Limits) are in Indata. 
   # Add them if not, make them numeric if they are
  if(any(names(Indata)=="MQL")){
    Indata$MQL<-as.numeric(Indata$MQL)} else {Indata$MQL<-as.numeric(NA)}
  
  if(any(names(Indata)=="UQL")){
    Indata$UQL<-as.numeric(Indata$UQL)} else {Indata$UQL<-as.numeric(NA)}

#### Create Data part of each characteristic ####
  if(wqx){
  MetaData$Data<-MetaData %>% dplyr::select(SiteCodeWQX, DataName) %>% 
    pmap(.f=function(SiteCodeWQX, DataName) {
    dplyr::filter(Indata, SiteCode == !!SiteCodeWQX, Characteristic == DataName) %>% 
              dplyr::select(-SiteCode,-Characteristic)})
  }
  
  if(!wqx){
    MetaData$Data<-MetaData %>% dplyr::select(SiteCode, DataName) %>% 
      pmap(.f=function(SiteCode, DataName) {
        dplyr::filter(Indata, SiteCode == !!SiteCode, Characteristic == DataName) %>% 
          dplyr::select(-SiteCode,-Characteristic)})
  }
  
#### Change numeric data to numeric, but the leave the rest as character ####
  NumDat<-MetaData$DataType=="numeric"
  MetaData[NumDat,]$Data<-suppressWarnings(MetaData[NumDat,]$Data %>% map(.f=function(x) mutate(x,Value = as.numeric(Value) ))) 

#### Create Characteristic objects ####
  MetaData$Characteristics<-MetaData %>% 
    dplyr::select(CharacteristicName, DisplayName, Units, Category, 
                  CategoryDisplay, LowerPoint, UpperPoint, LowerDescription, 
          UpperDescription, AssessmentDetails, Data) %>% 
    pmap(.f=new, Class="Characteristic")


#### Create a df of sites with correct Characteristic objects ####
  AllSites<-MetaData %>% group_by(ParkCode, SiteCode, SiteName,Lat,Long,Type) %>%
    summarize(Characteristics=list(Characteristics) %>% list) %>%
    ungroup

  
#### Create Park objects ####
   Parks<-MetaData %>% dplyr::select(Network, ParkCode, ShortName, LongName) %>% distinct() %>% 
    pmap(.f=new, Class="Park")

  
###### Make a list of sites with each park a nested list
  PSites<-map(Parks, function(Park){
    SiteDf<-filter(AllSites, ParkCode==getParkInfo(Park, info="ParkCode") ) %>% 
      dplyr::select(-ParkCode)
    SiteList<-pmap(.l=SiteDf, .f=new, Class="Site" )
  })
  
  
### Join Park objects with the sites  
  Parks<-map2(.x=Parks, .y=PSites, .f=function(x,y) {x@Sites<-y
  x} )
  
  return(Parks)
}