ESDL Gitlab Webpage:
We subscribe to open-source and reproducible science. Our group is now using Gitlab to distribute software resources. You can find our public repository here. Contents currently include:
- TransEnPackage: Matlab code for computing transfer entropy and mutual information between pairs of time series, together with significance thresholds. Gaps are ignored in the formulation of probability distributions.
- FuncConnectivityPackage: All software and data needed to formulate functional connectivity networks based on multiple linear regression. This package contains all of the code and data needed to reproduce the analyses in Larsen, L. G., S. Newman, C. Saunders, and J. W. Harvey. Complex networks of functional connectivity in an isolated wetland reconnected to its floodplain. Submitted to Water Resources Research, 2017.
- LargeScaleConnectivityPackage: Determines the functional form for properly upscaling surface-water flow resistance for flow through a heterogeneous landscape (i.e., floodplains, wetlands, deltas). This code relates landscape characteristics (directional connectivity, anisotropy, areal coverage) to the generalized average exponent in the upscaling relationship. The package was developed in support of Larsen, L., J. Ma, and D. Kaplan. How important is connectivity for surface-water fluxes? A generalized expression for flow through heterogeneous landscapes. Submitted to Geophysical Research Letters, January 2017.
Our directional connectivity software will also be available soon on Gitlab. For now, you can access it here…
Directional Connectivity Software
Here we provide the package of Matlab files needed to compute the directional connectivity index and connectivity-orientation curves, as described in Larsen et al., Ecological Applications, 2012. A readme in the zipfile describes the codes contained within the package.
To run this source code, Matlab is needed, together with the Matlab Image Processing toolbox and the MatlabBGL library. The latter is open source and can be downloaded here.
This modification to the method computes the directional connectivity index on a circular image, ensuring that the same image area is used in the computation as the image rotation is performed and avoiding boundary effects. The input is still a standard rectangular image, but users then have the opportunity to have the code inscribe or circumscribe a circle to crop the image to the area used for the computation. Circumscribing a circle would be more appropriate when running a computation on, for example, a pruned part of a river network; inscribing a circle would be more appropriate when running a computation on a landscape scale.
See the readme file within the supplement for instructions on getting started.