data-raw/01-data-process.R

In jmzobitz/periodicFitting: Fits periodicially repeating data (trigonometric, polynomial, and piecewise linear functions)

# AUTHOR: JMZ
# Modified: 1/1/19
# Script code to datasets - I am using Andy and Brandon's files


library(tidyverse)
library(lubridate)
library(devtools)

#

### DATA SET 1: MAUNA LOA CO2
# I downloaded the data and then modified it in Excel to strip trailing white space, makes it easier to read in.
mauna_loa <- read_csv('data-raw/co2_mm_mlo-jz.csv',na="-99.99")  %>%
  select(decimal_date,average) %>%
  rename(time=decimal_date,value=average) %>%
  mutate(date = date_decimal(time),product="co2") %>%
  na.omit() %>%
  mutate(date=floor_date(date,unit="month")) %>%
  select(date,time,product,value)
use_data(mauna_loa,overwrite = TRUE)

### DATA SET 2: CUMULUATIVE NEE FROM COLORADO
colorado <- read_csv('data-raw/niwotNEE.csv') %>%
  mutate(date=ymd(TIMESTAMP)) %>% # Convert to days
  rename(nee = NEE_CUT_REF) %>%
  mutate(time = decimal_date(date)) %>%
  filter(year(date) > 1998) %>%
  mutate(cumNEE=cumsum(nee)) %>%
  group_by(year=year(date)) %>%
  mutate(annNEE = cumsum(nee)) %>% ungroup() %>%
  select(date,time,annNEE,cumNEE) %>%
  gather(key=product,value=value,annNEE,cumNEE) %>% arrange(date)

use_data(colorado,overwrite = TRUE)

### DATA SET 3: GPP and ET from Nicaragua
et_data <- read_csv('data-raw/ET_1km_Farm_01_Area_579_meters(2018).csv') %>% select(year,month,ET_1km)

# Make the GPP data timestamps similar to the ET data
gpp_data <- read_csv('data-raw/Gpp_1km_Farm_01_Area_579_meters.csv') %>% select(year,month,Gpp_1km)


nicaragua <- et_data %>% left_join(gpp_data,by=c("year","month")) %>%
  mutate(date=date_decimal(year+month/12),time=year+month/12) %>% select(-year,-month) %>%
  gather(key=product,value=value,Gpp_1km,ET_1km) %>% arrange(date) %>%
  mutate(date=round_date(date,unit="day"))

use_data(nicaragua,overwrite = TRUE)