Author: Brett Bernstein

The code in runPlots.m (a Matlab file) 
randomly generates a sequence of spikes and
sample locations, and solves the recovery problem in 3 cases:
noiseless, dense noise, sparse noise.  The results are displayed in the
figures.  To changed from the Gaussian kernel to the Ricker, swap the
choice of kernel at the top of the file as commented.  All of the
other .m files are used by runPlots.m to generate data, run the
optimization, or generate plots.  In particular, tvMin.m,
tvMinNoise.m, and tvMinSparse.m contain all of the optimization code.
CVX is required.
