[DOC] Add "missing data" viz to example notebooks

[ericmjl]

This is quite important. I'd like to wrap other packages rather than invent the wheel. One possible package is to wrap missingno with a user-friendly API.

[souravsingh]

I am willing to work on this.

[ericmjl]

Thanks @souravsingh! Looking forward to the PR.

[JoshuaC3]

I assume there has been no PR for this? If so, I am happy to pick this up if still desired.

My only question is, how would this differ from just using missingno itself? What added functionality would we like to see? Thanks :)

[ericmjl]

@JoshuaC3 thanks for pinging in on this one! At the moment, I haven't had the bandwidth to give this more thought, so I'm very open to discussing the most appropriate use cases. What ideas do you have?

A few points I can think off the top of my head right now that might be relevant:

1. Janitor DataFrames inherit directly from Pandas dataframes and merely add in data cleaning convenience methods, so janitorDFs and pandasDFs are extremely compatible.
2. One thought that I ever once had was "just wrap missingno" --> the key thing that Janitor provides is method chaining, BUT method chaining might not be a relevant pattern for data sanity checks.

[JoshuaC3]

IMO, I prefer using the .pipe method on janitor function, as opposed to the jn.DataFrame with methods so this doesn't seem 100% natural to me. That said, I use the df.plot functionality a lot.

Also, missingno is already very easy to use,

import missingno as mn
mn.matrix(df)

This can be piped as well,

df.pipe(mn.matrix)

Thinking about this, I makes most sense to add this if we wish to extend functionality. I will have a think about how we could do this in janitor and post back.

[ericmjl]

This can be piped as well,
df.pipe(mn.matrix)

Oooh cool stuff! I didn't realize that, actually. Thanks for sharing this!

[ericmjl]

@JoshuaC3 I think this should be shown inside an example Jupyter notebook! What do you think? We have a bunch of notebooks already present, and can show how we go from data that is dirty (and has nullity) to data that is clean and densely populated!

[dancassin]

Once the PR #6 in pyjanitor-examples is accepted, then I believe this issue can be closed as well.

[ericmjl]

Thank you, @dancassin! 🎉