README.md

In andrewdolman/ecostattoolkit: An R package to implement and test the ECOSTAT toolkit for estimating nutrient boundaries

ecostattoolkit

An R package to implement and test the ECOSTAT toolkit for estimating nutrient boundaries.

Implements and tests the ECOSTAT toolkit for estimating nutrient boundaries. R functions have been created that replicate the original scripts by Geoff Phillips.

Installation

ecostattoolkit can be installed directly from GitHub using the install_github function in package devtools

# install.packages("devtools")
devtools::install_github("andrewdolman/ecostattoolkit")

Usage

Required packages

library(ecostattoolkit)
library(dplyr)
library(tidyr)

Read in and prepare data

filename <- system.file("extdata", "DataTemplate_Example1.csv", package = "ecostattoolkit")
dat <- read.csv(filename)

status_boundaries <- c("HG" = 0.8, "GM" = 0.6, "MP" = 0.4)

# min and max nutrient value used for model
nut.min <- 1    
nut.max <- 138  

dat$status <- cut(dat$EQR,
                  breaks = rev(c(Inf, status_boundaries, -Inf)),
                  labels = rev(c("High", "Good", "Moderate", "Poor")))

dat$x.u <- log10(dat$P)
dat$y.u <- dat$EQR

knitr::kable(head(dat))

| Record| Unique_ID| EQR| P| Exclude_P| N| Exclude_N| BioClass| Group | status | x.u| y.u| |-------:|-----------:|------:|-----:|-----------:|-----:|-----------:|---------:|:------|:---------|---------:|------:| | 1| 1| 0.781| 55.3| 0| 1.39| 0| 4| BE | Good | 1.742725| 0.781| | 2| 5| 0.771| 76.0| 0| 4.33| 0| 4| BE | Good | 1.880814| 0.771| | 3| 50| 0.505| 57.2| 0| 1.22| 0| 3| DE | Moderate | 1.757396| 0.505| | 4| 59| 1.015| 44.4| 0| 1.09| 0| 5| DE | High | 1.647383| 1.015| | 5| 60| 0.661| 28.6| 0| 0.64| 0| 4| DE | Good | 1.456366| 0.661| | 6| 63| 0.438| 93.3| 0| 1.19| 0| 3| DE | Moderate | 1.969882| 0.438|

Estimate boundaries with all methods

dat_sub <- filter(dat, Exclude_P == 0,
                  P <= nut.max,
                  P >= nut.min)

boundaries <- boundaries_all(dat_sub)


boundaries <- dplyr::mutate_if(boundaries, is.numeric, dplyr::funs(round(10^., 0))) %>% 
  select(-MP) 

knitr::kable(boundaries)

| Method | HG| GM| |:-----------|----:|----:| | OLS1_YonX | 27| 53| | OLS2_XonY | 30| 43| | RMA | 28| 49| | MM | 31| 45| | AdjQ | 25| 48| | Median | 30| 50| | q75 | 27| 53|

Compare with results from scripts in TKit_20170208_CORRECTED

Run script

source(system.file("extdata",
                   "TKit_20170208_CORRECTED/TKit_fit_lin_mod1.R",
                   package = "ecostattoolkit"))

# Collect results
script.res <- data.frame(
  Method = c(rep("OLS1_YonX", 2), rep("OLS2_XonY", 2),
             rep("RMA", 2)),

  Boundary = c("GM", "HG", "GM", "HG",
               "GM", "HG"),

  Value = c(GM.mod1, HG.mod1, GM.mod2, HG.mod2,
            GM.mod4, HG.mod4),
  stringsAsFactors = TRUE)

script.res %>% 
  tidyr::spread(Boundary, Value) %>% 
  dplyr::select(Method, HG, GM) %>% 
  arrange(Method) %>% 
  knitr::kable(.)

| Method | HG| GM| |:-----------|----:|----:| | OLS1_YonX | 27| 53| | OLS2_XonY | 30| 43| | RMA | 28| 49|

andrewdolman/ecostattoolkit documentation built on May 10, 2019, 10:30 a.m.