operation: central function for dataset implementation

[TirelessClock]

Written in context of #1319, DFFML could build a centralised module for implementation of datasets. That is, at present, in order to implement a new dataset, we need to write a new file for it, and deploy it on its own. Having familiarised myself with the iris dataset, I feel like the process is largely procedural, and we could work on functionalities to building one code for implementing any generic dataset. This would have long term implications, especially as it could help users drag and drop their own datasets.

Please mark this as an enhancement tag if approved @programmer290399 @pdxjohnny