The dask-ml statistic backend is available in pyXpcm but this library only offers the K-Means approach, since this is the most efficiently and easily scalable method for large dataset.
I raise an issue here, to mention that we gonna be working on implementing an optimized algorithm for the GMM method in dask_ml: dask/dask-ml#113
This also relates to the more general enhancement requirement reported here: #5
The dask-ml statistic backend is available in pyXpcm but this library only offers the K-Means approach, since this is the most efficiently and easily scalable method for large dataset.
I raise an issue here, to mention that we gonna be working on implementing an optimized algorithm for the GMM method in dask_ml: dask/dask-ml#113
This also relates to the more general enhancement requirement reported here: #5