inst/data_organization_workflow.R

In AgronomicForecastingLab/RS_PlantingDate: RS_PlantingDate

require(tidyverse)
require(RSPlantingDate)

# Organize the Beck's data and the extracted NDVI values

rm(list=ls())
set.seed(102396)
setwd('/Volumes/UIUC/libs/RS_PlantingDate')

# 1: Compile data from CSV files 
csvs = list.files('inst/data/NDVI/', full.names = T)
full = read.csv(csvs[1])
for (c in csvs[-1]){
  temp = read.csv(c)
  full = rbind(full, temp)
}
full = full %>% 
  mutate(Date = as.Date(substr(time, 1,10))) %>% 
  dplyr::select(ID, NDVI, Date)
rm(csvs,c,temp)

# 2: Load Beck's site information with sampled sites and updated coordinates
load('inst/data/finalSiteLocations.Rdata')
siteLoc = siteLoc %>% mutate(PLANTED = as.Date(PLANTED),
                             HARVESTED = as.Date(HARVESTED))
  
# 3: Add missing date values for planting date => for these values, Alex and I manually had to look at each PDF and extract the needed information
# mine = read.csv('inst/data/missingDates.csv') %>% 
#   mutate(PLANTED = as.Date(PLANTED, format = '%m/%d/%y'),
#          HARVESTED = as.Date(HARVESTED,format = '%m/%d/%y'))
# missing = c()
# for (i in 1:nrow(siteInfo)){
#   if (is.na(siteInfo$PLANTED[i]) | is.na(siteInfo$HARVESTED[i])){
#     X = which(mine$ID == siteInfo$ID[i])
#     
#     if (length(X) == 0){
#       missing = c(missing, i)
#       next
#     }
#     
#     siteInfo$PLANTED[i] = mine$PLANTED[X]
#     siteInfo$HARVESTED[i] = mine$HARVESTED[X]
#   }
#   
#   string = toString(siteInfo$PLANTED[i])
#   string2 = toString(siteInfo$HARVESTED[i])
#   siteInfo$PLANTED[i] = as.Date(paste0('2017-',substr(string, nchar(string)-4, nchar(string))))
#   siteInfo$HARVESTED[i] = as.Date(paste0('2017-',substr(string2, nchar(string2)-4, nchar(string2))))
# }

# 4: Update the latitude and longitude coordinates for this site 
# load('inst/data/updated_coordinates.Rdata')
# temp = left_join(siteInfo %>% dplyr::select(-Latitude, -Longitude),
#                  updated4 %>% dplyr::select(ID, Latitude, Longitude), by = c('ID'))

# 5: Combine information into one data frame
temp = left_join(full, siteLoc, by = 'ID') %>% 
  filter(!is.na(PLANTED)) %>% 
  mutate(#das = as.integer(Date - PLANTED), 
         dos = as.integer(PLANTED - as.Date(paste0(Year-1,'-12-31'))), 
         datYear = lubridate::year(Date)) %>%
  filter(datYear == Year) %>%
  dplyr::select(-datYear)

write.csv(temp, file = 'inst/data/organized/cleanedData.csv')