PluMA plugin that runs the Convergent Cross Mapping algorithm (CCM, Sugihara et al 2012) for detecting causality in complex ecosystems.
The plugin accepts as input time series data in a CSV file formatted in the following way:
"data1","data2", "timepoint1","value for data1","value for data2", "timepoint2","value for data1","value for data2", ...
The plugin will then output in CSV format a series of statistics that estimate, at every timepoint, the ability of data set 1 to predict values in data set 2. These include:
rho: The correlation coefficient between observed values in data set 2, and those predicted by the model mae: The mean absolute error between observed values in data set 2, and thos e predicted by the model rmse: The root mean square error between observed values in data set 2, and thos e predicted by the model
Note: example data (chicken_egg.csv) was assembled from the ChickEgg dataset of the lmtest library. That library is not required for this plugin, since we are using a CSV file.