The condensation method exploits the heterogeneity of the probability distribution functions (PDF) of event locations to reduce the size of seismic catalogs while preserving the overall spatial probability density. Ranked by decreasing location errors, the PDFs of events are successively condensed onto better located and lower variance event PDFs.
A data set containing the 1992 M7.3 Landers aftershocks is also provided for demonstration purposes.
Detailed description of the method can be is published in the following paper:
Y. Kamer, G. Ouillon, D. Sornette and J. Wössner (2014) Condensation of earthquake location distributions: Optimal spatial information encoding and application to multifractal analysis of South Californian seismicity http://arxiv.org/abs/1407.1036
Yavor Kamer (2020). Condensation of earthquake location distributions: Optimal spatial information encoding (https://www.mathworks.com/matlabcentral/fileexchange/48702-condensation-of-earthquake-location-distributions-optimal-spatial-information-encoding), MATLAB Central File Exchange. Retrieved .
Added the randnorm.m file from the Lightspeed library
Added 2D/3D ellipsoid plots using Gautam Vallabha's function
Corrected a bug which caused some the off-diagonal elements in the covariance matrices to be set to zero
added cover image
Inspired by: PLOT_GAUSSIAN_ELLIPSOID