Sheldon, you can edit your email header and footer setting before you send your posts, this is a standard one sent automatically. About benefits, people set their own. I also keep all my content free, many put paywalls for paid subscribers only.
No worries. I follow quite a few people on SubStack. Sometimes they have something juicy, like Terry Glavin, but only people that subscribe get that juicy stuff. It's how he set it up.
Hi Sheldon, Just saw your 220511 twitter thread re the funky AB death data. As another person who's consumed that data, and has a bit of a data background as well I reach many of the same conclusions you do - keep it up!. On tonight's data topic specifically, my hypothesis is that as the raw csv data only contains one field for date, it is most likely the 'reported_on' event, which may not correlate with the date of death.
Early on with that dataset I noted that AB was posting remarks about assumptions on the recovered date being assumed 14 days after confirmed (or reported on) date. So it stands likely that the date field does not represent the date of death, but rather the date the case was reported on.
As a death can occur on different days I suspect it'll almost impossible to corroborate precisely.
I've also seen the data set sort order change as well.
If you want, you can use the wayback machine to go back your 120 to January 9th and pull the csv that was scrapped on (or around) that day and compare some delta's.
Good luck and happy drinking data warrior adventures to you sir.
Sheldon, you can edit your email header and footer setting before you send your posts, this is a standard one sent automatically. About benefits, people set their own. I also keep all my content free, many put paywalls for paid subscribers only.
We Got You Sheldon! I Do Not Put A Price On Free Speech Either!
No worries. I follow quite a few people on SubStack. Sometimes they have something juicy, like Terry Glavin, but only people that subscribe get that juicy stuff. It's how he set it up.
Thank you for your work, and for sharing it (for free!) A man of integrity 😎
Hi Sheldon. I see you are not on LinkedIn. Is there another way to message you? I value your insights.
Hi Sheldon, Just saw your 220511 twitter thread re the funky AB death data. As another person who's consumed that data, and has a bit of a data background as well I reach many of the same conclusions you do - keep it up!. On tonight's data topic specifically, my hypothesis is that as the raw csv data only contains one field for date, it is most likely the 'reported_on' event, which may not correlate with the date of death.
Early on with that dataset I noted that AB was posting remarks about assumptions on the recovered date being assumed 14 days after confirmed (or reported on) date. So it stands likely that the date field does not represent the date of death, but rather the date the case was reported on.
As a death can occur on different days I suspect it'll almost impossible to corroborate precisely.
I've also seen the data set sort order change as well.
If you want, you can use the wayback machine to go back your 120 to January 9th and pull the csv that was scrapped on (or around) that day and compare some delta's.
Good luck and happy drinking data warrior adventures to you sir.