For a basic run with default parameters, load input data as a tensor X (tensortools sptensor, for example) and run 
MYRON(X, K, {}). The tensortools package is required.
where K is the desired group count.

For more detailed control of parameters, replace the third argument with a struct containing relevant fields

.la: regularization on sparsity of solution
.maxitr: maximum upper-level iterations
.maxtime: maximum time allowed for computation
.tol: convergence tolerance
.shape: 'bursty' or 'wavelet' or 'none' (the latter is a simple nonnegative constraint)
	.params: contains {.ab, .a0, .b, .g} for initializing bursty form parameters
	.thresh: wavelet coefficient truncation threshold
	.N: wavelet depth
.dense: if 1, expect a dense tensor (not generally recommended)

other tuning (and experimental) parameters are available

to run the included dataset (provided as a flat text file):


N = 100;
T = 750;
K = 2;
A = dlmread('synthetic.txt');
A(:,1:3) = A(:,1:3)+1;
X = sptensor(A(:,1:3),A(:,4))[N N T]);
ops.shape = 'bursty';
%Other parameters as required
res = MYRON(X,K,ops)


Please, cite:

@inbook{MYRON2021,
    author = {Gorovits, Alexander and Zhang, Lin and Gujral, Ekta and Papalexakis, Evangelos and Bogdanov, Petko},
    title = {Mining Bursty Groups from Interaction Data},
    year = {2021},
    isbn = {9781450384469},
    publisher = {Association for Computing Machinery},
    address = {New York, NY, USA},
    url = {https://doi.org/10.1145/3459637.3482370},
    booktitle = {Proceedings of the 30th ACM International Conference on Information & Knowledge Management},
    pages = {596–605},
    numpages = {10}
}

      



