a random statistics question
Uncategorized
3
Posts
2
Posters
0
Views
-
a random statistics question
let's say HYPOTHETICALLY that your group chat watches movies and there's three of you, and maybe you recently compiled a spreadsheet of everybody's movie rankings
stdev seems like the right tool to find most contentious/most agreed on movies for three people or pairwise between two people
is there a method to figure out which movies are ones where two people agreed and the other person disagreed?
-
undefined oblomov@sociale.network shared this topic
-
a random statistics question
let's say HYPOTHETICALLY that your group chat watches movies and there's three of you, and maybe you recently compiled a spreadsheet of everybody's movie rankings
stdev seems like the right tool to find most contentious/most agreed on movies for three people or pairwise between two people
is there a method to figure out which movies are ones where two people agreed and the other person disagreed?
@picklish variance?
When all three disagree, will be highest. When all three agree, variance will be lowest, and with only one disagreeing, variance will be intermediate.
-
@picklish variance?
When all three disagree, will be highest. When all three agree, variance will be lowest, and with only one disagreeing, variance will be intermediate.
@oblomov Yeah this all makes sense to me, and I can sort by variance to find the most/least agreement.
However, an intermediate variance could be either two people agreeing and one really disagreeing OR everybody slightly disagreeing? It feels harder to "find" those values, if that makes sense.
I guess I'm looking for some formula/function where a maximum value of the formula is one where two people agree and one disagrees. I'd even be fine if this was a specific per-person formula, like person A disagrees and people B and C agree and I just replicate that formula.
Like one hypothetical to find one where A disagrees is:
sqrt((B-A)^2 + (C-A)^2) - stdev(B, C)
essentially the stdev with A as the mean (how much B and C disagree with A) and then counteract that value by subtracting the stdev of B and C (how much B and C agree)
and this ~sort of works but also still has very high values in ways that don't make sense
mostly curious if I'm just making stuff up or there's something obvious I should consider
Feed RSS
Gli ultimi otto messaggi ricevuti dalla Federazione
-
@emama @attualita alla prossima commemorazione ci andranno tutti quelli infiltrati in askatasuna (organizzazione pacifista) che hanno messo a ferro e fuoco Torino...da capire da che parte staranno
-
-
@simonezanella e c'è anche il tram ristorante ottimo l'ho provato
-
@ClayAdventuresontheShelf @Madmonkey
Ma almeno abbiamo sortito qualche effetto pubblicitario?
Tradotto: qualcuno ti ha ordinato degli animaletti?
-
@stranobiovolta
Merde. Uno e una due.
Di cuore proprio.
@attualita
-
@odisseaquotidiana grazie adesso chiedo alla IA
-
Pause Print, Add Hardware, and Enjoy Strength
3D Printing is great, but it is pretty much the worst way to make any given part– except that every other technique you could use to make that part is too slow and/or expensive, making the 3D print the best option. If only the prints were stiffer, stronger, more durable! [JanTech Engineering] feels your plight and has been hacking away with the M601 command to try embedding different sorts of hardware into his prints for up to 10x greater strength, as seen in the video embedded below.
It’s kind of a no-brainer, isn’t it? If the plastic is the weak point, maybe we could reinforce the plastic. Most concrete you see these days has rebar in it, and fiber-reinforced plastic is the only way most people will use resin for structural applications. So, how about FDM? Our printers have that handy M601 “pause print” command built in. By creatively building voids into your parts that you can add stronger materials, you get the best of all possible worlds: the exact 3D printed shape you wanted, plus the stiffness of, say, a pulltruded carbon-fiber rod.
[JanTech] examines several possible inserts, including the aforementioned carbon rods. He takes a second look at urethane foam, which we recently examined, and compares it with less-crushable sand, which might be a good choice when strength-to-weight isn’t an issue. He doesn’t try concrete mix, but we’ve seen that before, too. Various metal shapes are suggested — there are all sorts of brackets and bolts and baubles that can fit into your prints depending on their size — but the carbon rods do come out ahead on strength-to-weight, to nobody’s surprise.
You could do a forged carbon part with a printed mold to get that carbon stiffness, sure, but that’s more work, and you’ve got to handle epoxy resins that some of us have become sensitized to. Carbon rods and tubes are cheap and safer to work with, though be careful cutting them.
Finally, he tries machining custom metal insets with his CNC machine. It’s an interesting technique that’s hugely customizable, but it does require you to have a decent CNC available, and, at that point, you might want to just machine the part. Still, it’s an interesting hybrid technique we haven’t seen before.
Shoving stuff into 3D-printed plastic to make it a better composite object is a great idea and a time-honored tradition. What do you put into your prints? We’d love to know, and so would [Jan]. Leave a comment and let us know.
