> Box plots [...] assume that your data follows a bell/gaussian shape.
Not sure how to square that with this statement on Wikipedia's page on box plots:
Box plots are non-parametric: they display variation in samples of a statistical population without making any assumptions of the underlying statistical distribution[3]
If you want to see why that is not fully correct you should read the article. For a box plot you need to calculate mean, variance and certain percentiles. These values don't make sense if your distribution does not follow a certain shape (because these values unambiguously define such a shape). See the examples in the article for what happens if you still try to use them in those cases. You can still extract the values of course (hence probably why wiki says they don't assume anything), but you lose significant information about the distribution. So you can no longer reverse the process.
I've never understood this to be the purpose of a boxplot, only a means of visualizing a distribution's quartiles.
You've gotten a flood of comments from upset people, so I'll keep it short by saying that a boxplot doesn't actually do what you claim for Gaussians, as the 0 and 100 percentile "whiskers" would be at plus/minus infinity. As for a bounded bell-shaped distribution, there are several non-unique ways to define such a distribution.
Which is why I find the article so compelling because I'd always read box plots as being about variance. To me the plot implied a quite normal distribution.
Sure. But if someone is using, for example, a notched boxplot to quickly evaluate differences in medians (i.e., they know how to correctly interpret a boxplot), it can still be a useful plot that conveys specific information that you would otherwise not get when looking at a violin plot, histogram, kernel density estimate or a strip plot.
My point, again, was: just because a boxplot is not useful to some people, doesn't mean that it is not a useful plot (particularly when augmented with a rugplot or a strip plot). Plots are not just used to convey information to others: they are also a useful tool in exploratory data analysis.
Notice that you can also apply the same critique to almost any plot: some people don't know how to interpret a violin plot (or kernel density estimate plot) correctly... does that make them useless?
The main advantage of a boxplot is that it is parameter-free (unlike histograms, violin plots and kernel density plots) and quickly conveys very specific information (median, range, quantiles, confidence interval for the median) that other types of plot usually don't.
Absolutely. A subset of the people in the audience are tenured or tenure-track academics. These folks get grants for postdocs. Moreover, they may be on a hiring committee. Even better: if you leave a favorable impression and a colleague of theirs has a job opening, they might recommend you.
Treat (academic) presentations as job talks. Good impressions open doors, not just in the immediate future but also further out than that.
> HN tends to be highly skeptical of “if you don’t have anything to hide, you wouldn’t mind the lack of privacy” arguments.
Well, those arguments flip the narrative. No one has to justify their privacy; invasion of privacy needs to be justified - and, when warranted, scope needs to be enforced.
In case anyone needs an analogue: when paying, you don't give the seller access to all of your money. They justify a particular amount and if you agree, they get the right to only that.
With money, we tend to call folks who take more than warranted "robbers" or "thieves". I think society would be in a better state of we viewed invasion of privacy as negatively.
> As long as the coverage and convenience are so lopsided,
Two ways to fix that (in general): improve one side, or worsen the other. Some measures do both simultaneously, eg. bus lanes take away asphalt for cars and improve bus infrastructure.
It's fairly easy to make a bus network more convenient than a car in a city. You can easily take measures such as dedicated bus lanes, one way streets, pedestrian areas, narrower streets + wider pavements and cycle lanes, less and more expensive parking.
The current attractiveness of public transport is roughly how attractive cars should have been. But most US cities have focused for decades on making car use more attractive. Congratulations, it worked.
The cost to redo what is there is so high that efforts for road diets are usually only a few blocks or a few miles at a time. More money need to be made available for these projects by the federal government as municipalities are usually strapped for cash.
That's a vast oversimplification. Try flipping it to anything outside your personal scope of skills and knowledge to see how wrong this position is.
Example: tap water quality (assuming you live where tap water is safe to drink). Do you know how it works? What steps are being taken? Could you fix that yourself for your house if things break down? And yet, you probably care.
Another example: car safety features. Could you add a crumple zone to an 80s car? A cage construction? Yet you probably care that any car you're in has those properly engineered and no part of your body will be crumpled in case of a collision.
Home batteries are a terrible idea. As a municipality -- or a neighbour -- you don't want 30 houses with 35 different batteries, installed by some nephew who's handy with DIY-stuff.
To just timeshift from daylight to evening/night/morning, you'd need batteries in the 25-35kWh range, 40-50kWh if airco on in the night, another 80kWh if you're driving electric and typically away from home during the day, charge at night. That is a lot of energy to pack, even at the low end. Imagine if everyone finds the cheapest set on AliExpress and DIYs it themselves, and then looks for a new model once current installation no longer matches their needs (capacity loss or use increase)... this must be a fire brigade worst nightmare.
Due to generation being more localised, we also need some form of much more local redistribution and/or storage. But not the nightmare that is one or more poorly installed and poorly maintained batteries of dubious origin per house.
There are a couple points to consider and a couple that refute some of your concerns.
First, you don’t need a full off grid battery setup to benefit. With time of use rates a battery that covers peak usage and even recharges under low time of use rates is a big benefit over its lifetime.
On safety, to sell back to the grid you need an interconnect agreement and inspection of your system. California utilities can install meters that can apparently detect secondary sources of back feeding the grid. Lastly if you try to DIY a correct setup you’ll have to buy UL1741 spec equipment for managing back feeding and equipment with other safety standards relevant to your use. Often this is not a code requirement but in the interconnect agreement and utilities have enforcement powers.
For off grid solar it’s a different story but the NEC always applies. If someone wants to do a comically bad job at their electric work they’re probably going to burn down their house anyway.
In my current home I’ve recently finished removing and replacing all the electric work accomplished by a previous owner’s “contractors.” Nothing like having electrical work run through low voltage grade audio visual wall plates and connectors for zero fire protection IN AN AIR PLENUM THAT GOES FROM MY BASEMENT TO MY SECOND FLOOR. That way if something does catch on fire it didn’t have all those pesky layers of protection between the floors of the house. Literally $30 in parts later and shorting wires get two hours to melt and trip a breaker.
The energy flowing through the batteries is the same despite the different battery brands and inverters. There's now APIs like Enode and others that have open standards for connecting dozens of EVs, batteries, HVAC equipment to react to grid events. Hopefully not too many DIYers out there but we're starting to see new models emerge where aggregators like Tesla, Sunrun, Currents, etc. (still needing PTO) can join the grid thanks to the APIs
Beyond fire risks or whatever, I actually don't see a problem with different battery models etc. The interface to the grid isnt exactly complicated, and decentralised storage seems like it would actually be quite robust and useful.
In a legal sense, most jurisdictions require some form of certification for such systems anyway, and the true diy versions are never going to be very common regardless.
And if you commision an artist to draw a black-leather tight-fit clad red-head superspy in an ad for tour product, it need not look like Black Widow from the MCU.
But if it does look very much like her, it doesn't really matter whether you never intended to.
The point isn't the time line of hiring the voice actor. The question is whether OpenAI was deliberately trying to make the voice sound like Johansson.
Suppose someone asked Dall-e for an image of Black Widow like in the first Avengers movie, promoting their brand. If they then use that in advertising, Johansson's portrait rights would likely be violated. Even (especially) if they never contacted her about doing the ad herself.
This is similar to that, but with voice, not portrait.
that's because one can make the argument that dall-e was regurgitating - it would be different if you get somebody who happens to look like her to pose in a similar way.
I don't think this is entirely right (not a lawyer)
You can't hire an artist to draw black widow in the style of Scarlett Johansson's widow. The issue isn't how the art is made, it's whether the end result looks like her.
I think there may be additional issues (to be determined either in courts or by Congress, in the US) with regard to how Dalle makes art, but elsewhere in the thread someone mentioned the Ford Bette middler case, and that does seem to be relevant (also, though, not exactly what happened here)
I don't have the expertise to know how similar this is to the case at hand.
Hmmms. While I definitely can see SO's arguments concerning deletion, that letter seems to blatantly contradict GDPR's right to be forgotten, which Wikipedia describes as a more limited "right to data erasure" [1].
To coin a Dutch phrase: I cannot make chocolate out of that. Anyone here have an idea how to bring these two points together? Other than the obvious "wrt. EU inhabitants, SO is lying", that is. Or is it really that simple?
Not sure how to square that with this statement on Wikipedia's page on box plots:
Box plots are non-parametric: they display variation in samples of a statistical population without making any assumptions of the underlying statistical distribution[3]