I'm doing a clandestine little project for Reddit karma, basically making a sort of infographic showing the top franchises subjected to R34 (primarily based on Paheal, but I wanted to use e621 and one or two other archives as alternate sources). Its pretty easy to just sort by most popular, but is there any record of tag counts increasing per month or anything? I know I saw someone make a graph like that a few years ago (I think also based on Paheal) but I wouldn't know how to get that information if its even available.

I don't know that any such information exists in an easily read format, but it's supposed to be possible to search for okami date:2018/06/01..2018/01/01 and see how many posts exist for that period of time. Either I'm getting the syntax wrong or the date search range doesn't actually work for some reason.

Furrin_Gok said:
I don't know that any such information exists in an easily read format, but it's supposed to be possible to search for okami date:2018/06/01..2018/01/01 and see how many posts exist for that period of time. Either I'm getting the syntax wrong or the date search range doesn't actually work for some reason.

You got it backwards. It's earliest date to latest date, like this: okami date:2018/01/01..2018/06/01

Furrin_Gok said:
I don't know that any such information exists in an easily read format, but it's supposed to be possible to search for okami date:2018/06/01..2018/01/01 and see how many posts exist for that period of time. Either I'm getting the syntax wrong or the date search range doesn't actually work for some reason.

Some working examples:
okami date:6_days_ago..20_days_ago
okami date:2018/01/01..2018/06/01

Though, what I'd probably do is hit up the e621 API for tags by count, retrieve a few pages' worth of that data, and then filter locally by tag type to whittle it down to copyrights and/or characters. (Do the requests once and stow the results, you can always re-parse your local copy later as you debug whatever you write to analyze the data.)

Then just start grabbing unfiltered post lists from the API, since the API includes the tagset for each returned post. Break them up into date ranges, and compare the tags against whatever you're looking for. Spit out a comma-separated list of date ranges and counts for each tag, open with your preferred spreadsheet software, make chart. (Again, grab the data once and do the parsing locally. No sense banging away at e6 needlessly while you debug your stuff.)

It may require a little coding savvy, but a project like this probably needs some coding savvy anyways, unless you intend to manually bang away at e6 and count the number of returned results by hand, which would be "eww" for any large-scale project.

archive.org has at least a few datapoints dating back to 2016, but generally only for the top 30 tags or so. With the right know-how, you could write a script to iterate through a month's posts, then count the results for a particular tag search. The *_yestermonths_ago notation would probably help with this, which would make dealing with e6's 320-post pagination the hardest part.

Yeah, basically do what the guy above me said.

BooruHitomi said:
You got it backwards. It's earliest date to latest date, like this: okami date:2018/01/01..2018/06/01

ikdind said:
Some working examples:
okami date:6_days_ago..20_days_ago
okami date:2018/01/01..2018/06/01

What the heck? "Ago" is sooner to earlier, but numerics are earlier to sooner? Somebody should update the search cheat sheet with that information.