[data cleaning] Mumbai getting repeated

[answerquest]

Describe the bug There are stray data entries where the city is "Mumbai City" or "Mumbai city", "Mumbai Suburb" etc instead of just Mumbai where the majority of entries are.
They are all showing up as separate clusters

To Reproduce Steps to reproduce the behavior:

1. Go to https://cluster.covid19india.org/
2. Click on City view
3. Pull Maharashtra State out to disentangle it and pull its cities out around it
4. See the multiple cities for Mumbai

Expected behavior One Mumbai

Screenshots

Desktop (please complete the following information):

• OS: Ubuntu 18.04
• Browser: Chromium
• Version: Version 80.0.3987.87 (Official Build)

Additional context

• Probably one or more of the data entry people thought about separating out parts of Mumbai.
• I'll advise against that, since we want to look at places that are well-connected. There isn't much separating Mumbai "city" from Mumbai "suburbs".
• And if you do want to split it up, then grouping all of Mumbai's suburbs together also doesn't make sense, since some suburbs are more disconnected from each other than from the main city (and what is the main city?)
• Creating one "city" for each suburb also doesn't make sense, as that will crowd out the actual cities in Maharashtra.
• If you do want to separate out, then please provision another grouping level below City. But I'm guessing that will involve coding changes too and may make this already-difficult-to-disentangle viz more complicated.

Redirection

• In case the data is being curated and maintained by other folks, kindly redirect me to them. Thank you for all the amazing work you're doing. This viz is extremely helpful.

[someshkar]

It's being curated by the covid19india ops team, can you bring this up on the Telegram group? You can find the link on the main website.

[answerquest]

@someshkar thanks for the direction. Telegram group seems to be locked most of the time now. Scaling up issues - seems their numbers exploded too.

[someshkar]

This is getting fixed in #88 Looks good? We're still looking for a general solution though.

[the-solipsist]

The "location" needed for cluster-mapping may differ from the "location" needed for other kinds of data analytics. A generalized solution for cluster-mapping be to use an additional field as "canonical cluster location" (e.g., "Mumbai"), which may be different from the actual location (e.g., "Navi Mumbai"). I would urge against discarding accurate data in the collaborative spreadsheet just to make it fit the needs of cluster-mapping.