Elsevier does not help science, and an argument can be made that they actually hinder it instead. Their journal subscription costs go up all the time, which means that universities have to spend more to keep their subscriptions and get access to new articles. They charge as much to publish as higher-impact journals, and they add an additional fee to make an article open access.
We should be spending research money on research, instead thanks to Elsevier we are spending research money on publication fees. They may have provided a useful service a decade (or more) ago when research universities were not as well interconnected and literature searches were more time consuming, but their current model does not fit with the modern state of scientific research. They need to adapt, or die. Keeping their current MO does not help anyone.
Why the fuck is that even allowed in any sane democracy ?
Because it is almost always executed on election day, when getting enforcement of voting laws is so insanely difficult to be almost a fruitless endeavour. One of the most common voter suppression techniques - which was executed very well in favor of GWB in 2004 - is to intentionally provide nowhere near enough voting machines in districts that you know you will lose. People can't take all day off to vote, and will at some point get out of line and go home (or to work). The distribution of voting machines in Ohio (and other places) pretty well matched the median income levels, rather than the population density.
Anything that discourages voting is voter suppression. The problem is that if you have been standing in line for 4 hours to vote you don't want to go spend several hours trying to find a voting official to lodge a complaint with. Indeed, voter suppression should never be tolerated. However in application is it just another election tool.
We have no way of knowing anything at that level of detail
True, though based on the situation you may be able to estimate some of them fairly well.
what percent of the population as a whole was infected
If you are the only hospital in the area, you can presume that all the severe cases are coming to you. You could then presume that a certain percent of infections are severe enough to warrant hospitalization and estimate the total infected from that. Obviously not a perfect number but a useful one nonetheless.
how many were vaccinated
This one also varies with the population pool you are dealing with. If you are the primary source of vaccination then you have a pretty good idea of how many are vaccinated. On the other hand if you have 20 clinics in your area, plus 4 drugs stores and 2 discount retailers that all do vaccination as well, then your numbers won't describe the vaccination rate well on their own.
The government does not have such information either
Correct. Of course, you do know who makes the vaccines, and you could use their numbers as a high limit for vaccinated people across the country.
What you describe is for hypothetical simulations.
Hypothetical situations are described to make models for the real world when real world data is not sufficient.
I'm just working with ER managers to enable efficient resource allocation:
Which is very important work as well. I'm just suggesting you may be able to get some reasonable estimates of the bigger picture from not a lot more data and work.
how many flu patients do we expect today? Where are we relative to seasonal baseline?
Also worth knowing. And with some additional work you should be able to project fairly well where you are on the trajectory, as well (unless, of course, you are one hospital in a community of millions, at which point the numbers likely break down).
In other words, I think what you describe is very interesting. I'm just suggesting that for many cases you may be able to use it to model some things that you had not (yet) described. I'm an informatics guy, I enjoy mining data like that...
I figured being you are a statistician you would make the necessary corrections, but it is easy to overlook some of the population adjustments you need to make for something like this. I guess you nee to decide whether or not the population of children who could potentially be infected is large enough in comparison to the spread of the virus to be considered truly constant, as of course no child should be infected more than once (assuming only one strain is passing through your area). Hence of course if the starting population was X, and Y were vaccinated, you then have a vulnerable population (X-Y), which should decrease over time as more people are infected and develop immunity.
Our of curiosity, what percent of the infected population are you assuming are being brought in to the hospital? That should figure in to your numbers and your model as well...
I am a statistician. In light of this year's early flu epidemic, I am tasked with modeling ER flu counts as a function of time.
When I plot the residual graph (observed - expected), I get upward spikes lasting about a week, corresponding to epidemics of particular strains. But there are also downward spikes lasting about a week. They occur at random, independent of the upward spikes. So what do I call such downward spikes? I've searched around but there is no antonym for "epidemic".
First of all, I don't think it's really fair that you were moderated off-topic, your comment is as close to topic as most. I quoted the full message so that more people will see it.
That said, I would first check the size of the population against the duration of the downward spikes. If the population is small enough relative to the size of the health care infrastructure you would expect the downward spikes to gradually get longer as the population builds immunity to the virus.
As for naming the downward spikes, I've never heard a standard name for it. Dwell, lag, or recovery might be reasonable? I presume those times don't represent a total of zero patients coming in, but rather far fewer patients, so it is only downward on a relative basis any ways...
The full text (available for free from anywhere - hooray for open access!)) states that the patient reported in June of this year. Paper was submitted on October 24, accepted November 1, and published November 20.
This also shows how good next-generation sequencing (NGS) technologies have become. They were able to sequence and assemble an entire virus genome in ~4 months or (likely) less, from a single infected human.
What do you mean by "decent linux graphics support"? I have a Thinkpad with NVidia NVS 3100M discrete graphics and 512mb vram. I'm perfectly content with it for what I do, which includes 3d molecular modelling. KDE looks great, too. On the other hand I don't play any 3d games so I can't tell you what Call of Duty 12 or any of those look like on here. I would sooner write code in CUDA for the GPU than do that. R
In other words, your sense of "decent linux graphics support" might not be the same as everyone else.
Students who want to avoid $200,000 in student-loan debt
If you are taking on $200,000 for a 4-year degree, you're doing it wrong. While it is increasingly more difficult every year to work your way through college (as I did), nobody should need to take on this much debt for a 4-year degree. Likely someone taking on that much debt is living way outside their means.
something more like this but without the $120 price tag
It seems to me that if it works, it would be $120 well spent. It would improve your parents' internet connection, which is your main goal, right? It would also require a minimal time investment on your behalf, which should also be a goal for you as you did not mention making frequent visits there. Third, if something goes wrong you have a manufacturer and a vendor to talk to - rather than having to rebuild the damned thing on your own.
One way to look at this is that if you visit once per year (you do see your parents once a year, right?) this will cost you only $10 per month. That is nothing. Besides, if you're three hours from the closest town, what are the chances you'll be able to get anything useful for building / repairing an antenna when you're out there? You seem to be about to start yourself on a project with very low probability of success for no apparent reason.
Buy the antenna and then spend your time building them a home file server to back up their important documents instead. Much better use of time and more tangible results.
Sure, the LHC generates tons of data. But that will soon be dwarfed by some of the data sets that are coming from proteomics research, especially when you take into account how many proteomics labs around the world are (or soon will be) capable of generating 1TB/day.
I don't know anyone who ever complained about the original Wii being "too big" (it's already the smallest home console on the market). Or, for that matter, anyone who has purchased one in the past 1-2 years and complained of it being "too expensive" (it's already the least expensive home console on the market). You can buy the regular (white) Wii for $120 right now at Best Buy, the $20 savings is not huge.
I suspect that these ones also have broken backwards-compatibility with the gamecube, so you are definitely getting less usefulness for less money.
What "threshold"? If pot really has such an impairment effect, then why not just use the standard sobriety test.
There is no impartial field (standard) sobriety test. It is up to the officer administering the test to both administer the test and interpret the results. It is very easy to skew or mangle this, on purpose or otherwise. When you have something that independently produces a number, you have an impartial result.
Every time I've been in a discussion on pot here on slashdot, this is what I have asked for - a standard and reproducible test for when someone has consumed too much. Strangely enough, it causes the pro-pot people to call me a fascist and the anti-drug people to call me a druggie.
I'm glad to know that there is so much room for a middle ground in slashdot political discussion...
There is still a significant market for DVD rental in this country - senior citizens. They tend to own TVs and DVD players but they tend not to own computers (or at least, with high-speed internet). Stock up on the movies they want to rent. Offer a senior discount (maybe just on Tuesdays or something). Advertise in areas where they go and places where they live.
Naturally, that market is time-limited, but you should be able to make money off of it for a while yet. While it's still viable you could even try running it as a senior-centric internet cafe, offering them assistance with things that might not be easy for older folks to do online.
Some people have been suggesting that Nintendo might pull out of the hardware (at least non-portable console) market entirely if the Wii U doesn't go over well. I am one who would very much not want to see that. Hopefully this console goes over well, even if it doesn't make the grand entrance that we saw with the original Wii.
And yes, I do still play my original Wii. I even bought a new game for it this week.
Most likely people have looked into this before, but I would suspect my blu-ray player is not running windows. Maybe instead of trying to get netflix to work in a browser in WINE in linux, they should try looking at how blu-ray players do it?
I rather doubt it would work out quite so well in Texas. Ultimately there are at least three different populations in Texas with regards to the border:
People who need the cheap labor (they don't want to secure it)
People who talk about it, but don't want to spend money on it (they won't secure it)
People who want to spend money on it, but won't do it themselves (they might secure it)
There may - though also may not - be a small population who is actually willing to do something themselves to secure the border. If any such people exist they are nowhere near numerous enough to be able to pull it off themselves.
Of course, the irony is that the border might end up becoming less appealing to immigrants since they wouldn't want to be in Texas anyways. Immigrants will instead pass through New Mexico, Arizona, and California because they won't want to be anywhere near the crazed Texans.
In other words, the Texans can look forward to paying a lot more for produce if they secede.
Texas has far more of the Mexican border than any other state. If they become an independent nation they will no longer get money from the US government to patrol and defend that border. On top of that, they'll then likely need to keep people from illegally leaving via the northern border...
The upshot is that competitors can now manufacture cheaper, generic versions of Viagra for sale in Canada.
I get email offers on a daily basis from "Canadian Pharmacy" offering to sell me generic Viagra - or Viagra with other ED medications in a "super pack" - all the time. You don't mean to say that some of the email offers I get aren't genuine, do you?
Actually, he just wants the government to "get out of marriage". Hence he supports gay marriage rights by not acknowledging any marriage rights.
is pro-choice
His official statement is something like "until the fetus is viable". Naturally that threshold will keep being adjusted down until eventually there is no more abortion.
anti-war
Again, you're not reading the full story, you're just taking your favorite sound bite. He doesn't want the US government going to war, because it costs money. However his platform will strip what little we have left of workers' rights, which will result in Texaco teaming up with United Defense and Coca-Cola to wage whatever war they like, with their employees as conscripted soldiers. Don't want to fight? Then no job for you!
drug legalization
They support that only to bring in tax revenue. Mainstream conservatives will eventually adopt that same idea for the same reason, as it allows them to place more taxes on people who make less money.
I fail to see how that makes him a conservative
I can't force you to actually read.
What crack are you smoking?
Are you going for some sort of irony award with that statement?
That's because you mispelled it. Rocky Anderson of the Justice Party.
Read the damned summary that slashdot posted. It says "Rock Anderson" not "Rocky Anderson". It's not my fault that slashdot cares so little about him as to not bother spelling his name correctly. I searched for the name provided by the summary.
Gary Johnson - Mostly a copy of Ron Paul, who is a republican who is so conservative that he makes Ronald Reagan look like Gandhi.
Jill Stein - Green party candidate, the only person (including Obama) in this election from any party who is less conservative than Ronald Reagan
Virgil Goode - Constitution party, another southern conservative who has convinced himself that every problem this country faces comes down to there being too many people running around here who don't look and sound like he does
Rock Anderson - A candidate so obscure he doesn't even have a wikipedia entry. Being as even CmdrTaco has a wikipedia entry, that likely says enough right there. Similarly searching on google for "Rock Anderson" mostly tries to link to you the wedding of Kid Rock and Pamela Anderson; the first result might be the correct Rock Anderson but the page returns a database error instead. If you want to cast a protest vote you might do better by writing in Mickey Mouse.
Did they use a botnet that was scattered all over the world, or just a specific set of systems? I would recommend going through your logs to see what you can find out about the attack, there may be some patterns there that you can learn from.
That said, a lot of people suggest you contact the authorities. I would suggest that those people have probably never tried that themselves. The authorities - local or federal - generally don't give a shit about cyber crime. They give it some (virtual) lip service on their websites but when presented with actual cyber crime they always find something more interesting to do with their time. After all, you said the criminal was in Lebanon, and the FBI has no jurisdiction there. Even if you found an FBI agent who cared, he wouldn't be able to get interpol working on it before the (electronic) check is cashed and the culprit has cleaned up his tracks.
In other words, you have to do the work yourself. Maybe you can learn something from the logs, or maybe you'll need to look at distributed hosting to better prepare yourself for a potential future attack.
Slashdot Mirror
Elsevier does not help science, and an argument can be made that they actually hinder it instead. Their journal subscription costs go up all the time, which means that universities have to spend more to keep their subscriptions and get access to new articles. They charge as much to publish as higher-impact journals, and they add an additional fee to make an article open access.
We should be spending research money on research, instead thanks to Elsevier we are spending research money on publication fees. They may have provided a useful service a decade (or more) ago when research universities were not as well interconnected and literature searches were more time consuming, but their current model does not fit with the modern state of scientific research. They need to adapt, or die. Keeping their current MO does not help anyone.
Why the fuck is that even allowed in any sane democracy ?
Because it is almost always executed on election day, when getting enforcement of voting laws is so insanely difficult to be almost a fruitless endeavour. One of the most common voter suppression techniques - which was executed very well in favor of GWB in 2004 - is to intentionally provide nowhere near enough voting machines in districts that you know you will lose. People can't take all day off to vote, and will at some point get out of line and go home (or to work). The distribution of voting machines in Ohio (and other places) pretty well matched the median income levels, rather than the population density.
Anything that discourages voting is voter suppression. The problem is that if you have been standing in line for 4 hours to vote you don't want to go spend several hours trying to find a voting official to lodge a complaint with. Indeed, voter suppression should never be tolerated. However in application is it just another election tool.
We have no way of knowing anything at that level of detail
True, though based on the situation you may be able to estimate some of them fairly well.
what percent of the population as a whole was infected
If you are the only hospital in the area, you can presume that all the severe cases are coming to you. You could then presume that a certain percent of infections are severe enough to warrant hospitalization and estimate the total infected from that. Obviously not a perfect number but a useful one nonetheless.
how many were vaccinated
This one also varies with the population pool you are dealing with. If you are the primary source of vaccination then you have a pretty good idea of how many are vaccinated. On the other hand if you have 20 clinics in your area, plus 4 drugs stores and 2 discount retailers that all do vaccination as well, then your numbers won't describe the vaccination rate well on their own.
The government does not have such information either
Correct. Of course, you do know who makes the vaccines, and you could use their numbers as a high limit for vaccinated people across the country.
What you describe is for hypothetical simulations.
Hypothetical situations are described to make models for the real world when real world data is not sufficient.
I'm just working with ER managers to enable efficient resource allocation:
Which is very important work as well. I'm just suggesting you may be able to get some reasonable estimates of the bigger picture from not a lot more data and work.
how many flu patients do we expect today? Where are we relative to seasonal baseline?
Also worth knowing. And with some additional work you should be able to project fairly well where you are on the trajectory, as well (unless, of course, you are one hospital in a community of millions, at which point the numbers likely break down).
In other words, I think what you describe is very interesting. I'm just suggesting that for many cases you may be able to use it to model some things that you had not (yet) described. I'm an informatics guy, I enjoy mining data like that...
I figured being you are a statistician you would make the necessary corrections, but it is easy to overlook some of the population adjustments you need to make for something like this. I guess you nee to decide whether or not the population of children who could potentially be infected is large enough in comparison to the spread of the virus to be considered truly constant, as of course no child should be infected more than once (assuming only one strain is passing through your area). Hence of course if the starting population was X, and Y were vaccinated, you then have a vulnerable population (X-Y), which should decrease over time as more people are infected and develop immunity.
Our of curiosity, what percent of the infected population are you assuming are being brought in to the hospital? That should figure in to your numbers and your model as well...
I am a statistician. In light of this year's early flu epidemic, I am tasked with modeling ER flu counts as a function of time.
When I plot the residual graph (observed - expected), I get upward spikes lasting about a week, corresponding to epidemics of particular strains. But there are also downward spikes lasting about a week. They occur at random, independent of the upward spikes. So what do I call such downward spikes? I've searched around but there is no antonym for "epidemic".
First of all, I don't think it's really fair that you were moderated off-topic, your comment is as close to topic as most. I quoted the full message so that more people will see it.
That said, I would first check the size of the population against the duration of the downward spikes. If the population is small enough relative to the size of the health care infrastructure you would expect the downward spikes to gradually get longer as the population builds immunity to the virus.
As for naming the downward spikes, I've never heard a standard name for it. Dwell, lag, or recovery might be reasonable? I presume those times don't represent a total of zero patients coming in, but rather far fewer patients, so it is only downward on a relative basis any ways...
The full text (available for free from anywhere - hooray for open access!)) states that the patient reported in June of this year. Paper was submitted on October 24, accepted November 1, and published November 20.
This also shows how good next-generation sequencing (NGS) technologies have become. They were able to sequence and assemble an entire virus genome in ~4 months or (likely) less, from a single infected human.
What do you mean by "decent linux graphics support"? I have a Thinkpad with NVidia NVS 3100M discrete graphics and 512mb vram. I'm perfectly content with it for what I do, which includes 3d molecular modelling. KDE looks great, too. On the other hand I don't play any 3d games so I can't tell you what Call of Duty 12 or any of those look like on here. I would sooner write code in CUDA for the GPU than do that.
R In other words, your sense of "decent linux graphics support" might not be the same as everyone else.
The oldest dinosaur? A Middle Triassic dinosauriform from Tanzania
The open access model would like to say "you're welcome".
Students who want to avoid $200,000 in student-loan debt
If you are taking on $200,000 for a 4-year degree, you're doing it wrong. While it is increasingly more difficult every year to work your way through college (as I did), nobody should need to take on this much debt for a 4-year degree. Likely someone taking on that much debt is living way outside their means.
something more like this but without the $120 price tag
It seems to me that if it works, it would be $120 well spent. It would improve your parents' internet connection, which is your main goal, right? It would also require a minimal time investment on your behalf, which should also be a goal for you as you did not mention making frequent visits there. Third, if something goes wrong you have a manufacturer and a vendor to talk to - rather than having to rebuild the damned thing on your own.
One way to look at this is that if you visit once per year (you do see your parents once a year, right?) this will cost you only $10 per month. That is nothing. Besides, if you're three hours from the closest town, what are the chances you'll be able to get anything useful for building / repairing an antenna when you're out there? You seem to be about to start yourself on a project with very low probability of success for no apparent reason.
Buy the antenna and then spend your time building them a home file server to back up their important documents instead. Much better use of time and more tangible results.
Sure, the LHC generates tons of data. But that will soon be dwarfed by some of the data sets that are coming from proteomics research, especially when you take into account how many proteomics labs around the world are (or soon will be) capable of generating 1TB /day.
I don't know anyone who ever complained about the original Wii being "too big" (it's already the smallest home console on the market). Or, for that matter, anyone who has purchased one in the past 1-2 years and complained of it being "too expensive" (it's already the least expensive home console on the market). You can buy the regular (white) Wii for $120 right now at Best Buy, the $20 savings is not huge.
I suspect that these ones also have broken backwards-compatibility with the gamecube, so you are definitely getting less usefulness for less money.
What "threshold"? If pot really has such an impairment effect, then why not just use the standard sobriety test.
There is no impartial field (standard) sobriety test. It is up to the officer administering the test to both administer the test and interpret the results. It is very easy to skew or mangle this, on purpose or otherwise. When you have something that independently produces a number, you have an impartial result.
Every time I've been in a discussion on pot here on slashdot, this is what I have asked for - a standard and reproducible test for when someone has consumed too much. Strangely enough, it causes the pro-pot people to call me a fascist and the anti-drug people to call me a druggie.
I'm glad to know that there is so much room for a middle ground in slashdot political discussion...
There is still a significant market for DVD rental in this country - senior citizens. They tend to own TVs and DVD players but they tend not to own computers (or at least, with high-speed internet). Stock up on the movies they want to rent. Offer a senior discount (maybe just on Tuesdays or something). Advertise in areas where they go and places where they live.
Naturally, that market is time-limited, but you should be able to make money off of it for a while yet. While it's still viable you could even try running it as a senior-centric internet cafe, offering them assistance with things that might not be easy for older folks to do online.
Some people have been suggesting that Nintendo might pull out of the hardware (at least non-portable console) market entirely if the Wii U doesn't go over well. I am one who would very much not want to see that. Hopefully this console goes over well, even if it doesn't make the grand entrance that we saw with the original Wii.
And yes, I do still play my original Wii. I even bought a new game for it this week.
Most likely people have looked into this before, but I would suspect my blu-ray player is not running windows. Maybe instead of trying to get netflix to work in a browser in WINE in linux, they should try looking at how blu-ray players do it?
There may - though also may not - be a small population who is actually willing to do something themselves to secure the border. If any such people exist they are nowhere near numerous enough to be able to pull it off themselves.
Of course, the irony is that the border might end up becoming less appealing to immigrants since they wouldn't want to be in Texas anyways. Immigrants will instead pass through New Mexico, Arizona, and California because they won't want to be anywhere near the crazed Texans.
In other words, the Texans can look forward to paying a lot more for produce if they secede.
Texas has far more of the Mexican border than any other state. If they become an independent nation they will no longer get money from the US government to patrol and defend that border. On top of that, they'll then likely need to keep people from illegally leaving via the northern border...
The upshot is that competitors can now manufacture cheaper, generic versions of Viagra for sale in Canada.
I get email offers on a daily basis from "Canadian Pharmacy" offering to sell me generic Viagra - or Viagra with other ED medications in a "super pack" - all the time. You don't mean to say that some of the email offers I get aren't genuine, do you?
Gary Johnson supports gay marriage rights
Actually, he just wants the government to "get out of marriage". Hence he supports gay marriage rights by not acknowledging any marriage rights.
is pro-choice
His official statement is something like "until the fetus is viable". Naturally that threshold will keep being adjusted down until eventually there is no more abortion.
anti-war
Again, you're not reading the full story, you're just taking your favorite sound bite. He doesn't want the US government going to war, because it costs money. However his platform will strip what little we have left of workers' rights, which will result in Texaco teaming up with United Defense and Coca-Cola to wage whatever war they like, with their employees as conscripted soldiers. Don't want to fight? Then no job for you!
drug legalization
They support that only to bring in tax revenue. Mainstream conservatives will eventually adopt that same idea for the same reason, as it allows them to place more taxes on people who make less money.
I fail to see how that makes him a conservative
I can't force you to actually read.
What crack are you smoking?
Are you going for some sort of irony award with that statement?
That's because you mispelled it. Rocky Anderson of the Justice Party.
Read the damned summary that slashdot posted. It says "Rock Anderson" not "Rocky Anderson". It's not my fault that slashdot cares so little about him as to not bother spelling his name correctly. I searched for the name provided by the summary.
Did they use a botnet that was scattered all over the world, or just a specific set of systems? I would recommend going through your logs to see what you can find out about the attack, there may be some patterns there that you can learn from.
That said, a lot of people suggest you contact the authorities. I would suggest that those people have probably never tried that themselves. The authorities - local or federal - generally don't give a shit about cyber crime. They give it some (virtual) lip service on their websites but when presented with actual cyber crime they always find something more interesting to do with their time. After all, you said the criminal was in Lebanon, and the FBI has no jurisdiction there. Even if you found an FBI agent who cared, he wouldn't be able to get interpol working on it before the (electronic) check is cashed and the culprit has cleaned up his tracks.
In other words, you have to do the work yourself. Maybe you can learn something from the logs, or maybe you'll need to look at distributed hosting to better prepare yourself for a potential future attack.
I've got a finger gesture for that.