In jiho/coclimer: Analyse Time Series of HABs for project CoClime

coclimer

R package to standardise statistical analyses of time series in the CoClime project.

Installation

The package is not on CRAN. Install it with

# if needed
install.packages("remotes")
# then
remotes::install_github("jiho/coclimer")

Usage

Load the package

library("coclimer")

A test dataset is included in the package, containing the concentration of Ostreopsis ovata in two forms and environemental variables. Read more about it with

?ost

Make it available and inspect it with

data(ost)
head(ost)

Your data should be made to look the same: a date column, columns for species concentrations/abundances, columns for environmental variables.

Functions plot_multi() and plot_seasonnal() allow to represent the data of multiple series graphically

# full time series
plot_multi(ost, benthic:planktonic)
plot_multi(ost, benthic:planktonic, trans="sqrt")
# seasonal view
plot_seasonal(ost, benthic:planktonic, trans="sqrt")

The function yearly_stats() allows to compute standardised statistics for each year.

# for benthic concentration
yearly_stats(ost$date, ost$benthic, bloom_threshold=200000)

To regress abundances of HAB-forming organisms on environmental variables, use the function relate_env(). This computes a quantile-based regression of abundances on all environmental variables using the Random Forest algorithm, computes partial dependence plots for the most relevant variables and plots them.

suppressMessages(library("tidyverse"))
d <- filter(ost, benthic>0) %>% as.data.frame()
relate_env(sqrt(d$benthic), select(d, chla:temperature), n=3)

Read the help of each function for more information.

jiho/coclimer documentation built on July 26, 2020, 4:02 a.m.