Roadmap for 1.0.0

[WenheLI]

This PR is to discuss what we should cover for our initial release.
First of all, a huge thank to @steveoni and @risenW for their work.

Right now, we have implemented basic operations for NLP and CV.
And we should discuss, what we should cover for our first release (Ideally, the end of this month; )

IMO, we should cover all functionalities in danfo, plus what covered in #8.
In this case, we will have the ability to cope with tabular data, visual data, and text.

Besides, we should have a more robust unit test system and comprehensive CI/CD for delivery.

Finally, we should add some benchmarks to indicate our performance and pressure-test for a large set of data.

[risenW]

Thanks @WenheLI I have been reading occupied on my end, and have not been able much. But with a planned released from month end, I'll definitely schedule more time on this

[yorkie]

What's the rest works for our first v1 release? For me, it's the utils to translate the JS tensor to Python/NumPy's.

/cc @WenheLI @risenW @steveoni

[yorkie]

And documentation is also required IMO.