|
|
# ? Apr 17, 2020 18:52 |
|
|
# ? May 19, 2024 18:43 |
|
Looks like I might have a publisher for my paper in progress: "Habitat Choice in Defecatory Preferences in American Black Bears: Woods vs. Others". I'll have to look for a social sciences publisher for my other paper: "Religious Sect Determination of Selected Popes"
|
# ? Apr 17, 2020 19:58 |
|
From Raoult's "study" on chloroquine/azithromycin, courtesy of Derek Lowe. Look at those dual Y axis scales https://blogs.sciencemag.org/pipeline/archives/2020/03/29/more-on-cloroquine-azithromycin-and-on-dr-raoult (And don't read the comments.)
|
# ? Apr 17, 2020 20:44 |
|
im probably trap springing myself here but i never figured out how to read those dumb double y axis charts
|
# ? Apr 17, 2020 21:06 |
|
Suspicious Dish posted:im probably trap springing myself here but i never figured out how to read those dumb double y axis charts Some of the lines/bars have their value to the right, some have their value to the left. Sometimes used to save space, some of the time intentionally misleading.
|
# ? Apr 17, 2020 21:15 |
|
Whenever you see multiple Y axes, some of the components on the graph are read on one and some on the other. They should be independent. In this case, the bar charts are on the left axis and represent total number of patients, the lines are on the right axis and represent percentage of the original patients. So they're representing the same data on both axes, which... I guess you can do it, but why? But the real stupidity comes in when you look at the scale of the right axis. The neat way of showing this IMO would be to scale both axes together so that 0 patients is 0% and 80 patients (the original number) is 100%. Stick them both on the left, and that could be a useful addition. But instead they're treated as independent variables, and at a first glance it looks like the cohort started with ~105% of patients and ended up with -10%. Then there's the poly fit, which serves no value at all. Who cares what the R^2 is? You're not constructing a model of the data for predictive purposes, the fit doesn't matter.
|
# ? Apr 17, 2020 21:20 |
|
so its a gimmick format intended to show correlations of independent variables, with a dece amount of fuckery sneaking into the axes. got it
|
# ? Apr 17, 2020 21:27 |
|
Karia posted:Then there's the poly fit, which serves no value at all. Who cares what the R^2 is? You're not constructing a model of the data for predictive purposes, the fit doesn't matter. R^2 is of absolutely no use whatsoever, particularly for predictive purposes.
|
# ? Apr 17, 2020 21:37 |
|
Suspicious Dish posted:so its a gimmick format intended to show correlations of independent variables, with a dece amount of fuckery sneaking into the axes. got it It can make sense when the two variables have different units, or different magnitudes of the same unit. But it gives the creator additional room to mess with the scaling and truncating of axes. Always be suspicious of graphs, be extra suspicious of graphs with two y axes.
|
# ? Apr 17, 2020 21:38 |
|
Platystemon posted:
|
# ? Apr 19, 2020 02:20 |
|
https://arxiv.org/abs/2004.10796
|
# ? Apr 24, 2020 22:12 |
|
Oh, a representative example.
|
# ? Apr 24, 2020 22:28 |
|
most of the other examples seem they took a bunch of movie stills and had people say what they thought about the two people interacting in the scene, then their NN rolled some dice and spit out some of the common options (ie. they wanted to talk, or they did not want to talk; possibly both or neither)
|
# ? Apr 24, 2020 22:50 |
|
I dunno, maybe he held it just because he likes how bronze feels. Could be a fetish thing.
|
# ? Apr 25, 2020 00:39 |
|
Kennel posted:I dunno, maybe he held it just because he likes how bronze feels. Could be a fetish thing. He does look ecstatic.
|
# ? Apr 25, 2020 01:57 |
|
Carthag Tuek posted:most of the other examples seem they took a bunch of movie stills and had people say what they thought about the two people interacting in the scene, then their NN rolled some dice and spit out some of the common options (ie. they wanted to talk, or they did not want to talk; possibly both or neither) there's a recent paper which trained one of these photo recognition/conversation networks which has state of the art performance, and also doesn't see any of the photos, it's 100% cold reading the user
|
# ? Apr 25, 2020 03:52 |
|
Tunicate posted:there's a recent paper which trained one of these photo recognition/conversation networks which has state of the art performance, and also doesn't see any of the photos, it's 100% cold reading the user Honestly this is probably easier to do than to actually build photo recognition software.
|
# ? Apr 25, 2020 04:31 |
|
The Cheshire Cat posted:Honestly this is probably easier to do than to actually build photo recognition software. extremely easier
|
# ? Apr 25, 2020 06:47 |
|
|
# ? Apr 25, 2020 09:03 |
|
Sorry about the huge text, I just like the chart range.
|
# ? Apr 25, 2020 10:37 |
|
What, you think "more" per 100k isn't clear?
|
# ? Apr 25, 2020 13:42 |
|
“Per 100k” is for cowards who are afraid of negative exponents, and it’s even worse when the legend isn’t quantified.
|
# ? Apr 25, 2020 13:49 |
|
What's the point of a state flag that you can't jack it to
|
# ? Apr 25, 2020 14:32 |
|
Phlegmish posted:What's the point of a state flag that you can't jack it to Luckily I have an anchor fetish
|
# ? Apr 25, 2020 14:39 |
|
Phlegmish posted:What's the point of a state flag that you can't jack it to You can jack it to any state flag if you try hard and believe in yourself.
|
# ? Apr 25, 2020 14:40 |
|
The Bloop posted:Luckily I have an anchor fetish Hell yeah~ I used to work right across the street from the Capitol so I got the giant anchor flags AND a good view of The Independent Man's toes. I nominate for the thread all the COVID-19 cumulative case/death graphs that fit a smooth curve to their data points and end up with days with negative deaths.
|
# ? Apr 25, 2020 15:21 |
|
Cichlidae posted:death graphs that fit a smooth curve to their data points and end up with days with negative deaths. Jesus was a statistical anomoly.
|
# ? Apr 25, 2020 15:45 |
|
You’re the expert.
|
# ? Apr 25, 2020 15:48 |
|
https://twitter.com/theonion/status/1234903384207110145?lang=en
|
# ? Apr 25, 2020 17:52 |
|
Phlegmish posted:What's the point of a state flag that you can't jack it to My state has a bare boob. She's standing on a dead guy, but w/e - BOOB!
|
# ? Apr 26, 2020 02:04 |
|
Phlegmish posted:What's the point of a state flag that you can't jack it to Buddy, they won't even let me jack off to it.
|
# ? Apr 26, 2020 02:11 |
|
Memento posted:Buddy, they won't even let me jack off to it. common misconception. flag code states you can't sexual gratify yourself to the national flag. state flags are fair game, with the important caveat that it's treason to bust to any confederate or confederate-derived flag.
|
# ? Apr 26, 2020 02:50 |
|
Mr. Fix It posted:common misconception. flag code states you can't sexual gratify yourself to the national flag. state flags are fair game, with the important caveat that it's treason to bust to any confederate or confederate-derived flag. waiting for the tweets "its a heritage boner not a hate boner"
|
# ? Apr 26, 2020 03:11 |
|
Actually it was about states rights to have boners
|
# ? Apr 26, 2020 03:15 |
|
Very few of the men who fought for the South actually owned boners
|
# ? Apr 26, 2020 04:22 |
|
Is this what the South will rise again is talking about?
|
# ? Apr 26, 2020 04:32 |
|
Idaho's flag's got a sexy lady and a sexy miner dude so there's more for everyone. Esto Perpetua? More like Estrus Perpetua.
|
# ? Apr 26, 2020 07:14 |
|
Mr. Fix It posted:common misconception. flag code states you can't sexual gratify yourself to the national flag. state flags are fair game, with the important caveat that it's treason to bust to any confederate or confederate-derived flag. The 2001–2003 flag of Georgia is bad, but is it Confederate-derived?
|
# ? Apr 26, 2020 07:55 |
|
Platystemon posted:The 2001–2003 flag of Georgia is bad, but is it Confederate-derived? It contains the 1956-2001 flag so yes
|
# ? Apr 26, 2020 08:32 |
|
|
# ? May 19, 2024 18:43 |
|
Platystemon posted:The 2001–2003 flag of Georgia is bad, but is it Confederate-derived? yeah, but it still is: first flag of the confederacy: georgia:
|
# ? Apr 27, 2020 01:01 |