Quantifying Ecological Memory in Palaeoecological Datasets and Other Long Time-Series
Align and join multiple time series to a common temporal resolution
Quantifies ecological memory with Random Forest.
Turns the outcome of runExperiment into a long table.
Extracts ecological memory features from the output of computeMemory...
Create lagged versions of time series variables
Plots the output of runExperiment.
Plots output of computeMemory
Computes ecological memory patterns on simulated pollen curves produce...
Quantifies ecological memory in long time-series using Random Forest models ('Benito', 'Gil-Romera', and 'Birks' 2019 <doi:10.1111/ecog.04772>) fitted with 'ranger' (Wright and Ziegler 2017 <doi:10.18637/jss.v077.i01>). Ecological memory is assessed by modeling a response variable as a function of lagged predictors, distinguishing endogenous memory (lagged response) from exogenous memory (lagged environmental drivers). Designed for palaeoecological datasets and simulated pollen curves from 'virtualPollen', but applicable to any long time-series with environmental drivers and a biotic response.