data-raw/soilsensors/code_proc-soilsensors.R

In vanichols/maRsden: Provides easy access to data collected from the Marsden Farm at Iowa State University

########################
#
# created: march 13 2020
#
# updated:
#
#
# purpose: process soil sensor data
#
# NOTES:
#
#
#########################


rm(list=ls())
library(tidyverse)
library(lubridate)
library(janitor) #--used to clean data

#--this is the plotkey data available in the package
pk <- read_csv("data-raw/plotkey/plotkey.csv") %>%
  filter(year %in% c(2018, 2019),
         harv_crop %in% c("C2", "C4"))



# raw data (from facts datahub) -------------------------------------------

ssraw19 <- readRDS("data-raw/soilsensors/tidyQC-soil-marsden.rds") %>%
  filter(year == 2019) %>%
  mutate(date = as_date(date),
         year = year(date),
         doy = yday(date),
         plot = as.numeric(plot))

ssraw18 <- readRDS("data-raw/soilsensors/2020-03-13mars-intervention.rds") %>%
  filter(year == 2018) %>%
  mutate(date = as_date(date),
         year = year(date),
         doy = yday(date),
         plot = as.numeric(plot))

ssraw <-
  ssraw18 %>%
  bind_rows(ssraw19)


ss <-
  ssraw %>%
  left_join(pk) %>%
  select(year, date, doy, plot_id,
         sensor_type, sensor_name, sensor_depth_cm, sensor_unit, value)


#--viz check
#--these probably need scaled somehow....
ss %>%
  filter(sensor_unit == "soilVWC") %>%
  left_join(pk) %>%
  ggplot(aes(doy, value, group = plot_id)) +
  geom_line(aes(color = rot_trt)) +
  facet_grid(sensor_depth_cm~year)

ss %>%
  filter(sensor_unit == "soilVWC") %>%
  left_join(pk) %>%
  group_by(doy, date, rot_trt, year, sensor_depth_cm) %>%
  summarise(value = mean(value, na.rm = T)) %>%
  ggplot(aes(doy, value, group = rot_trt)) +
  geom_line(aes(color = rot_trt)) +
  facet_grid(sensor_depth_cm~year)


ss %>%
  filter(sensor_unit == "wtdepth_mm", doy < 300) %>%
  left_join(pk) %>%
  ggplot(aes(doy, value, group = plot_id)) +
  geom_line(aes(color = plot_id)) +
  scale_y_reverse(limits = c(3000, 0)) +
  facet_grid(.~year)


# make datasets -------------------------------------------

mrs_soilsensors <-
  ss %>%
  arrange(year, doy, plot_id, sensor_depth_cm)

mrs_soilsensors %>%  write_csv("data-raw/soilsensors/soilsensors.csv")

usethis::use_data(mrs_soilsensors, overwrite = T)