Develop Process - same functionality as USIDataset

[ssomnath]

1. What changes would you like to make to this class if you could - this is your chance.
2. __init__ can't be taken more or less as is -
   • would need some way to figure out how to flatten the data into two axes - examples vs features (over which data would be processed)
3. Take these functions as is:
   • assign job indices
   • estimate compute time per pixel
   • get pixels in current batch
   • test
   • use_partial_Computation
   • set_cores
   • set_memory_and_cores
   • map_function
   • write_results_chunk
   • create_results_dataset
   • write_source_dset_provenance
   • get_existing_datasets
   • _unit_computation
4. Rewrite the following:
   • check for duplicates
   • set_memory
   • read_data_chunk
   • create_compute_status_dataset
5. Carefully rewrite compute()

[ssomnath]

Process class will expect NSIDataset to arise from HDF5 (data should exist safely on disk)

[ssomnath]

We will use @gduscher 's problem - Gaussian mixtures model applied to denoising EELS spectra as the basis for developing the new Process class. I am still not entirely sure where this class will live. There may be a possibility to generalize the Process class by finding commonalities with pyUSID.Process

[ssomnath]

This issue may need to be moved to sidpy