tensor_coclustering

CoClust_nD_Synth:

Create synthetic tensors and run tauTCC a specified number of times on each tensor.

Parameters:

1. n_test: integer
2. dimensions: list of tuples, separated by '-'. Ex: [100,100,30-50,50,50] is a list of the shape of the tensors
3. clusters: list of tuples, separated by '-'. Ex: [2,2,2-3,3,3-5,3,2] is a list of the combinations of number of classes on each mode; each tuple t (representing the number of clusters) must have len(t) == min([len(k) for k in dimensions]). When the the number of modes of a tensor is greater then len(t), the number of clusters on the last mode of t is repeated on all the missing modes. Ex: n_modes = 4, t= (5,3,2) ----> t = (5,3,2,2)
4. noise: integer.

output: a file with the results of the tests. The file will be saved in a new folder ./output

CoClust_3D_DBLP:

Run tauTCC on the DBLP Four-Areas dataset. Parameters (optional):

1. algorithm: string. Optional. Accepted values: ALT, AVG, AGG, ALT2, AGG2. Default: ALT2
2. level: string. Optional.Accepted values: DEBUG, INFO, WARNING, ERROR, CRITICAL. Default value: WARNING

CoClust_3D_MovieLens:

Run tauTCC on MovieLens datasets. Parameters:

1. tensor: integer. Accepted values: 1 or 2.
2. algorithm: string. Optional. Accepted values: ALT, AVG, AGG, ALT2, AGG2. Default: ALT2
3. level: string. Optional.Accepted values: DEBUG, INFO, WARNING, ERROR, CRITICAL. Default value: WARNING

CoClust_3D_yelp:

Run tauTCC on yelp datasets. Parameters:

1. tensor: integer. Accepted values: TOR or PGH.
2. algorithm: string. Optional. Accepted values: ALT, AVG, AGG, ALT2, AGG2. Default: ALT2
3. level: string. Optional.Accepted values: DEBUG, INFO, WARNING, ERROR, CRITICAL. Default value: WARNING

#######################################################################################

The data are in folder ./resources

The algorithm and the code for generate synthetic tensors are in folder ./algorithms