User talk:TiagoLubiana/Structured Discussions Archive 1

Hi Tiago - Thanks so much for inviting me to the WikiProject call. Unfortunately, I was preparing for a new course I just started today. Would you be available for another call some time this week or would you be open to exchanging emails? I love the idea of documenting more hospital information. I also know there's been some debate about storing data on tables on Wikipedia vs Wikidata. Thanks so much for your quick response!

Hello, Will,

I can imagine, lots of things happening. Sure, if you want to, thusrday is a good day for me for a call. Anytime would be fineǃ

You can also send me a mail and we can chat there (tiago.lubiana.alves̊ [at] usp.br, whatever suits you the best.

Best,

Tiago

Given your work on updating case count statements automatically with a bot, I was wondering if you'd be interested in helping to automate counts about U.S. counties from sources other than Johns Hopkins.

For weeks, this table of cases in California by county has been updated manually and piecemeal by various users. I'm not sure if it's been updated consistently. As a nod to automation, I've converted a couple rows to use tabular data, but the underlying tables so far need to be updated manually using a script. I have a Toolforge account and can develop a bot to update those tables, but managing all 55 counties, each with their own data table, would be a bit daunting.

For counties where Wikipedia isn't presenting a time series chart, I think it would suffice to keep a Wikidata statement up-to-date with the latest numbers. The Corona Data Scraper project should make it feasible to obtain current numbers for all the counties. Then it's a matter of feeding the numbers into Wikidata via bot. I could look into that eventually, but I figure you're further along and I'd rather not duplicate efforts.

I'm interested to hear your thoughts about where else we could streamline these mini data projects as well.

Hey, thanks for the messageǃ I am actually a bit overloaded .

I am not familiar of how to add tabular data on commons. Most of the work on this case count bot has actually been done by @jvcavv on this GitHub pageː https://github.com/lubianat/wikidata_covid19/tree/master/sandbox/worldwide_data/src . Maybe he is in a better position to aid this project.

Moving the data from the corona scraper to Wikidata seems straight forward, but I would not know how to do it for the .tab items.

Hey, sorry for taking so long to answer, I'm actually a bit overworked too.

I'm also not familiar on adding tabular data to commons, but if you intend on creating quickstatements to update the items on wikidata instead, I believe the script we're using to batch update case/death counts for countries (The one Tiago linked above) can be adapted to take in data coming from Corona DataScraper. That seems pretty straight forward and by the end you'd have a text file containing the quickstatements to do the batch update, but if you need any help feel free to hit me up.

Thank you for works to improve items related to Q84263196. But your recent contributions to Q84055514 makes double entry for same date and have some errors. Kindly check on this. Please note: Updation of this items are carried out on the data from https://www.mohfw.gov.in/

Hello, Gnoeee. Thank you for your notice.

I have added India to the list of country items that I will not update, great that you are working on that. Do you have an automation script?

Why do you say it has some errors? The data is from https://datahub.io/core/covid-19, which is a CC0 version of JHU portal.

Conflicting information should not be a problem on Wikidata, as long as the sources are correct, that is why ranking exists. Even if the information is wrong. If you think that editions of this item should be restricted to some format, this can be discussed further in depth.

But of course, you are maintaining and using the item, and likely https://www.mohfw.gov.in/ is a gold standard for India, so that is great, and I totally respect your decision. I will not change it anymore.

Hi, TiagoLubiana

We are currently doing daily updates Manually. Just using some scripts for copying references / creating duplicate items..

Yes the data from has a huge set of good data available under CC0. I agree with that point. But i have seen it have some difference in the Values for the same date.

Example case: has different Values in P8010 for same qualifier P585. So while fetching the data to Wikipedia or external sites the values gets summed.

And in English Wikipedia also follows the data from the same website for updation. That's why we would like to go up with the data available under https://www.mohfw.gov.in/

Was there a midday chat? It didn't end up on my calendar ~

Oh, sorry! I should have send a reminder.

Yes, there was one, you can find the discussion in this etherpad:

https://etherpad.wikimedia.org/p/open_meeting_COVID-19_20%2F04

Aha -- thank you!

Hi, I am watching your Lab at YouTube, https://www.youtube.com/watch?v=FX3eLxpEuwc Thanks! 8-)