In kimberly-bastille/ecopull: What the Package Does (One Line, Title Case)

knitr::opts_chunk$set(echo = TRUE,
                      message = FALSE)
devtools::load_all()
`%>%` <- magrittr::`%>%`

Rationale

I went back to our code to calculate updated indicators for 2023 and thought... there has to be a better way. The code was too long and complicated for what should be a fairly simple task of calculating seasonal OISST anomalies.

New seasonal OISST function

The new function make_seasonal_oisst takes a raster file name as input. The raster file should have a layer with gridded temperature data for each day. It reads in the file and loops through EPU and season to calculate the mean temperature for each EPU x season. The return value is a tibble.

Code

make_seasonal_oisst

New workflow

Code

readLines(here::here("_targets.R")) %>%
  cat(sep = "\n")

Comparison to existing OISST indicators

sst_anomaly %>%
  dplyr::mutate(Var = dplyr::case_when(stringr::str_detect(Var, "winter") ~ "Winter OISST anomaly",
                                       stringr::str_detect(Var, "spring") ~ "Spring OISST anomaly",
                                       stringr::str_detect(Var, "summer") ~ "Summer OISST anomaly",
                                       stringr::str_detect(Var, "fall") ~ "Fall OISST anomaly"),
                Type = "New method") %>%
  dplyr::rename(Time = Year) %>%
  dplyr::full_join(ecodata::seasonal_oisst_anom %>%
                     dplyr::mutate(Type = "Old method")) %>%
  dplyr::mutate(Var = Var %>%
                  stringr::str_remove(" OISST anomaly")) %>%
  ggplot2::ggplot(ggplot2::aes(x = Time,
                               y = Value,
                               color = Type)) +
  ggplot2::geom_point() +
  ggplot2::geom_line() +
  ggplot2::facet_grid(rows = ggplot2::vars(Var),
                      cols = ggplot2::vars(EPU)) +
  ggplot2::theme_bw()

sst_anomaly %>%
  dplyr::mutate(Var = dplyr::case_when(stringr::str_detect(Var, "winter") ~ "Winter OISST anomaly",
                                       stringr::str_detect(Var, "spring") ~ "Spring OISST anomaly",
                                       stringr::str_detect(Var, "summer") ~ "Summer OISST anomaly",
                                       stringr::str_detect(Var, "fall") ~ "Fall OISST anomaly")) %>%
  dplyr::rename(Time = Year,
                New_value = Value) %>%
  dplyr::full_join(ecodata::seasonal_oisst_anom) %>%
  dplyr::mutate(Var = Var %>%
                  stringr::str_remove(" OISST anomaly")) %>%
  ggplot2::ggplot(ggplot2::aes(x = Time,
                               y = New_value - Value)) +
  ggplot2::geom_col(color = "black",
                    fill = "maroon") +
  ggplot2::facet_grid(rows = ggplot2::vars(Var),
                      cols = ggplot2::vars(EPU),
                      scales = "free") +
  ggplot2::theme_bw()

OISST anomalies with new long-term mean

new_sst_anomaly %>%
  dplyr::mutate(Type = "New LTM") %>%
  dplyr::full_join(sst_anomaly %>%
                     dplyr::mutate(Type = "Old LTM")) %>%
  dplyr::mutate(Var = dplyr::case_when(stringr::str_detect(Var, "winter") ~ "Winter OISST anomaly",
                                       stringr::str_detect(Var, "spring") ~ "Spring OISST anomaly",
                                       stringr::str_detect(Var, "summer") ~ "Summer OISST anomaly",
                                       stringr::str_detect(Var, "fall") ~ "Fall OISST anomaly")) %>%
  dplyr::rename(Time = Year) %>%
  dplyr::mutate(Var = Var %>%
                  stringr::str_remove(" OISST anomaly")) %>%
  ggplot2::ggplot(ggplot2::aes(x = Time,
                               y = Value,
                               color = Type)) +
  ggplot2::geom_point() +
  ggplot2::geom_line() +
  ggplot2::facet_grid(rows = ggplot2::vars(Var),
                      cols = ggplot2::vars(EPU)) +
  ggplot2::theme_bw()