True, Matlab has included a just-in-time compiler, but the language has limitations and the compiler isn't as mature as Julia's (LLVM). Devectorized code often runs _faster_ in Julia, sometimes much faster. In Julia, vectorization is just a way of invoking BLAS functions when appropriate - as it should be.
Julia is more related to the lisp family of programming languages than ML. Some have said that Julia is just another lisp (albeit with a nifty dynamic type system supporting type families and polymorphism).
The best is actually faster than C, and the worst is 2.1x slower, not 5x slower (numerical values given in http://julialang.org/ ). And that's just the fibonacci benchmark; the other ones are within 50% of C performance except for rand_mat_stat which is 60% slower than C (this is due to the garbage collector, apparently).
I found that MATLAB's emphasis on vectorization forces you into a form of coding that is not always useful or productive. Sure, it's better to use a*b when multiplying matrices, but stuff like this? :
If you think Java handles 'all tasks' then Julia isn't aimed at you. it's for scientific/numerical computing. Java is terrible at that, in terms of syntax, library support, and speed. Even python is far better than Java for that purpose.
Not at all. Julia is just-in-time compiled to native code. You can basically get C-like performance in pure julia: http://julialang.org/benchmark...
There's also a powerful type system and lisp-style macros, along with support for parallel programming and lightweight threads, allowing you to do stuff like: https://gist.github.com/anj1/2...
CO2 sequestration is a "look over there!" solution. The energy costs in capturing, compressing, and transporting CO2 are huge, and where do you put the CO2 anyway? Few rock formations are suitable for permanent storage of CO2, and those that are suitable are rarely located near coal plants. As far as I know there's only one carbon capture operation running, and that one is only commercially viable because it's conveniently located near a bunch of oil wells that need the CO2 to push oil up out of the ground.
I honestly think fusion has a better chance of becoming viable in fifty years.
In comparison, gas is squeaky clean. Less CO2, virtually no fly ash, etc.
> I thought they were able to capture the emissions at these big plants.
Nope. Carbon capture and storage (CCS) is still a long ways off from being practical, if it can ever become practical. They can capture fly ash but where does it go? They usually just dump it all in a big pile beside the power plant. You can find tailings piles near most coal power plants. Water seeps into the piles and causes seepage of all sorts of nasty stuff like mercury and lead and radioactive ash into water supplies. About 95% of the mercury in the tuna you eat has come from burning coal. The FDA advises against pregnant women eating tuna for this reason.
Our diets weren't as diverse as vegetarians like to think. If you look at the diets of most major human populations, there are some common themes: A mix of gathered plant matter, including roots, leaves, fruit, and nuts. And hunted meat including insects, birds, and large game. Even Inuit ate lots of plant food, contrary to the widely-circulated myth that they just ate meat (an important part of the Inuit diet is the half-digested plant matter found in the stomachs of wild game). But meat seems to be a constant. You can not find a major human population that didn't eat meat. They all did. They HAD to, because paleolithic plants were not nearly as nutritious or calorie-rich as modern agricultural inventions like potatoes and corn. None of these modern agricultural marvels existed for our ancestors. But they had buffalo.
The average caloric mix seems to be about 70% plants and 30% meat.
You're only partially right. You CAN get all the protein you need from plants, but it's hard, and you need to eat a lot of plants, and you need to occasionally supplement with eggs or milk or otherwise consume processed plant material (like soy protein powder). It's not a 'natural' way of eating by any stretch of the imagination.
Bugs, on the other hand, are probably the most natural food for humans. Our ancestors consumed insects in abundance, and many still do. Insects formed large and important parts of our ancestral diets. We ate larvae, we ate roaches, we ate locusts. They're widely available, the protein content is perfect, the harmful fat and carbohydrate content is very low, and there are few insect diseases that are readily transmittable to humans. It's just that as we became more agricultural and started consuming tasty lamb and beef and chicken, insects became associated with the poor and unclean, and then gradually got forgotten entirely. But now we live in a world where we know that these animals aren't as good for us as insects are, and we're starting to hit severe limits on how many chicken and cattle we can raise. To expand the human population - or hell, even maintain it where it is - we might need to seriously consider eating insects again.
The porosity in question is different, more like zeolite than a liquid with bubbles in it. Plus, the liquid in question has permanent porosity, unlike soda (or any liquid with bubbles in it) which has only temporary porosity.
No, because his comment is brainless and moronic. He (or she or whatever) is questioning the use of words like 'toxic' and 'problematic' while at the same time asking the authors to prove that something is 'bad'.
If puritans actually _followed_ their religions instead of cherry-picking what they like, they would find that their religions expressly and clearly tell them to spread wealth as much as possible instead of accumulating it. And it's not like it's in the fine print -- there are 2000 verses in the bible dealing with poverty and social justice (in comparison, there are only six dealing with homosexuality). And this doesn't just apply to Christianity, it applies to most major religions. Islam for instance has the khoms and the zakaat - giving 1/5 of your wealth to the needy as a rule plus additional charities for special purposes.
I don't think it's necessarily a bad idea if done correctly. If you put in some rules so that the risk of maiming or injury goes down (no weapons, no below the belt, stopping when the other person says "STOP!!", etc.) then you get boxing/mma/wrestling, and that's perfectly legal. The problem is, you're always going to have the moron who shows up drunk, smashes the other guy's head in with a bottle, and gets the entire thing banned. It wouldn't even take a day. In the mean time, you can always join a legal martial arts or boxing class.
Most of what you're saying is untrue. Randomly picking a few points:
> Meanwhile gas cars are getting so efficient with their mileage;
Total bullshit. Gasoline engine efficiency hasn't increased much since the 80's. Of course car manufacturers would LIKE you to believe that their cars are now so much more efficient and green and all that, but it's marketing bullshit.
Not that there haven't been some small incremental improvements in efficiency. There have. But nothing revolutionary - just a few percentage points.
And there are just simple physical limits to gasoline car efficiency. No gas car engine is going to hit the efficiency of, say, a large gas turbine (over 35%) in typical driving conditions. It just can't be done in a car engine. The power/weight, acceleration, reliability/maintenance, and thermodynamic concerns are all working against it.
> and a more supportive infrastructure to help with repairs and maintenance
The model S's drive train has something like sixteen moving parts. And virtually no frictional contact except a few bearings. Compare with the hundreds (thousands?) of moving parts in a car engine, all furiously rubbing against each other.
Electric cars don't need maintenance, except for tires and brake pads. Motor failures are possible but very rare. The battery wears out gracefully and can be replaced, but doesn't really need it in practice. Under conditions of extreme stress the motor inverter might fail but even that is a freak scenario.
The kind of intensive and continuous maintenance you need with a gas car (oil changes, oil filters, air filters, antifreeze, transmissions, belts failing, radiators bursting, etc.) just don't exist in an electric car. It's just battery -> inverter -> motor -> gearbox -> wheel. An electric car is about as mechanically complicated as the engine _starter_ on a gas car.
> If we're just going to shove the problem of fossil fuel burning off on someone else (i.e. a power plant), we're not actually solving any environmental,
Yes we are. A large, stationary power plant is always going to be way more energy efficient than a small car engine. A car engine is about 20% efficient; new power plant designs are capable of nearly 80% efficiency (especially when used for combined heat and power). 60% efficiency is typical for even older designs. If you crunch the numbers, driving an electric car powered solely by coal-based electricity puts out about half as much carbon as a gas car. This includes all the inefficiencies in power line transmission, batteries, etc. In reality this figure is actually much better than that since not all electricity comes from coal. But the point is that even if we were to use the worst CO2 emitter - coal - we'd still be making an improvement over gasoline cars, and it would be an economic improvement as well since the price of coal is far more stable than oil. Of course the goal is to do even better than that - to power cars using nuclear, solar, wind, and other clean energy sources.
> Right now electric cars are financially supported by the state, while gas cars pay for the roads
Instead of whining about tax on gasoline for road maintenance, car owners should be thankful they aren't being taxed for carbon emissions, which would be the case in a fair world.
Citation needed. Official cost estimates aren't known but Tesla probably pays around $20,000 for the battery on the P85D, and that's a big battery. http://insideevs.com/tesla-bat...
> My current car is 20 years old and requires very little in terms of operating expenses.
Cool anecdote bro. The actual data reveals that a 10 year old mid-price-range car typically requires $500/yr in maintenance costs; a 20 year old car requires $1000/yr on average.
Slashdot Mirror
http://cyrille.rossant.net/wha...
> In Python those operations can be done with not-slow code with the proper library.
Not always.
Fact is, Python has a two-language problem.
True, Matlab has included a just-in-time compiler, but the language has limitations and the compiler isn't as mature as Julia's (LLVM). Devectorized code often runs _faster_ in Julia, sometimes much faster. In Julia, vectorization is just a way of invoking BLAS functions when appropriate - as it should be.
Julia is more related to the lisp family of programming languages than ML. Some have said that Julia is just another lisp (albeit with a nifty dynamic type system supporting type families and polymorphism).
import Base.UTF8proc.category_code
category_code('')
The best is actually faster than C, and the worst is 2.1x slower, not 5x slower (numerical values given in http://julialang.org/ ). And that's just the fibonacci benchmark; the other ones are within 50% of C performance except for rand_mat_stat which is 60% slower than C (this is due to the garbage collector, apparently).
I found that MATLAB's emphasis on vectorization forces you into a form of coding that is not always useful or productive. Sure, it's better to use a*b when multiplying matrices, but stuff like this? :
cell2mat(arrayfun(@(K) accumarray(C, F(:,K), [], @mean), 1:size(F,2), 'Uniform', 0))
come on. It pays to have a language where you can write out a clear explicit loop if you want to and not worry about performance.
http://julialang.org/benchmark...
If you think Java handles 'all tasks' then Julia isn't aimed at you. it's for scientific/numerical computing. Java is terrible at that, in terms of syntax, library support, and speed. Even python is far better than Java for that purpose.
How does that change the fact that Python is slow? Julia uses BLAS as a linear algebra backend too, but that has nothing to do with this.
Not at all. Julia is just-in-time compiled to native code. You can basically get C-like performance in pure julia: http://julialang.org/benchmark...
There's also a powerful type system and lisp-style macros, along with support for parallel programming and lightweight threads, allowing you to do stuff like: https://gist.github.com/anj1/2...
(that's just a toy example, of course)
CO2 sequestration is a "look over there!" solution. The energy costs in capturing, compressing, and transporting CO2 are huge, and where do you put the CO2 anyway? Few rock formations are suitable for permanent storage of CO2, and those that are suitable are rarely located near coal plants. As far as I know there's only one carbon capture operation running, and that one is only commercially viable because it's conveniently located near a bunch of oil wells that need the CO2 to push oil up out of the ground.
I honestly think fusion has a better chance of becoming viable in fifty years.
> Are modern coal plants really all that harmful to the environment?
Yes, they really are all that harmful to the environment. https://en.wikipedia.org/wiki/...
In comparison, gas is squeaky clean. Less CO2, virtually no fly ash, etc.
> I thought they were able to capture the emissions at these big plants.
Nope. Carbon capture and storage (CCS) is still a long ways off from being practical, if it can ever become practical. They can capture fly ash but where does it go? They usually just dump it all in a big pile beside the power plant. You can find tailings piles near most coal power plants. Water seeps into the piles and causes seepage of all sorts of nasty stuff like mercury and lead and radioactive ash into water supplies. About 95% of the mercury in the tuna you eat has come from burning coal. The FDA advises against pregnant women eating tuna for this reason.
I could go on.
Our diets weren't as diverse as vegetarians like to think. If you look at the diets of most major human populations, there are some common themes: A mix of gathered plant matter, including roots, leaves, fruit, and nuts. And hunted meat including insects, birds, and large game. Even Inuit ate lots of plant food, contrary to the widely-circulated myth that they just ate meat (an important part of the Inuit diet is the half-digested plant matter found in the stomachs of wild game). But meat seems to be a constant. You can not find a major human population that didn't eat meat. They all did. They HAD to, because paleolithic plants were not nearly as nutritious or calorie-rich as modern agricultural inventions like potatoes and corn. None of these modern agricultural marvels existed for our ancestors. But they had buffalo.
The average caloric mix seems to be about 70% plants and 30% meat.
That's why I said harmful fat, not all fat. E.g. trans fats, saturated fats. Learn to read, my friend.
You're only partially right. You CAN get all the protein you need from plants, but it's hard, and you need to eat a lot of plants, and you need to occasionally supplement with eggs or milk or otherwise consume processed plant material (like soy protein powder). It's not a 'natural' way of eating by any stretch of the imagination.
Bugs, on the other hand, are probably the most natural food for humans. Our ancestors consumed insects in abundance, and many still do. Insects formed large and important parts of our ancestral diets. We ate larvae, we ate roaches, we ate locusts. They're widely available, the protein content is perfect, the harmful fat and carbohydrate content is very low, and there are few insect diseases that are readily transmittable to humans. It's just that as we became more agricultural and started consuming tasty lamb and beef and chicken, insects became associated with the poor and unclean, and then gradually got forgotten entirely. But now we live in a world where we know that these animals aren't as good for us as insects are, and we're starting to hit severe limits on how many chicken and cattle we can raise. To expand the human population - or hell, even maintain it where it is - we might need to seriously consider eating insects again.
I've had a few bugs. They weren't so bad.
The porosity in question is different, more like zeolite than a liquid with bubbles in it. Plus, the liquid in question has permanent porosity, unlike soda (or any liquid with bubbles in it) which has only temporary porosity.
Figures that this would be the first response on slashdot.
Your two-bit brain is not capable of understanding the real world. Ask your parents for a genetic refund.
No, because his comment is brainless and moronic. He (or she or whatever) is questioning the use of words like 'toxic' and 'problematic' while at the same time asking the authors to prove that something is 'bad'.
You are a good demonstration of the problem.
If puritans actually _followed_ their religions instead of cherry-picking what they like, they would find that their religions expressly and clearly tell them to spread wealth as much as possible instead of accumulating it. And it's not like it's in the fine print -- there are 2000 verses in the bible dealing with poverty and social justice (in comparison, there are only six dealing with homosexuality). And this doesn't just apply to Christianity, it applies to most major religions. Islam for instance has the khoms and the zakaat - giving 1/5 of your wealth to the needy as a rule plus additional charities for special purposes.
I don't think it's necessarily a bad idea if done correctly. If you put in some rules so that the risk of maiming or injury goes down (no weapons, no below the belt, stopping when the other person says "STOP!!", etc.) then you get boxing/mma/wrestling, and that's perfectly legal. The problem is, you're always going to have the moron who shows up drunk, smashes the other guy's head in with a bottle, and gets the entire thing banned. It wouldn't even take a day. In the mean time, you can always join a legal martial arts or boxing class.
Most of what you're saying is untrue. Randomly picking a few points:
> Meanwhile gas cars are getting so efficient with their mileage;
Total bullshit. Gasoline engine efficiency hasn't increased much since the 80's. Of course car manufacturers would LIKE you to believe that their cars are now so much more efficient and green and all that, but it's marketing bullshit.
Not that there haven't been some small incremental improvements in efficiency. There have. But nothing revolutionary - just a few percentage points.
And there are just simple physical limits to gasoline car efficiency. No gas car engine is going to hit the efficiency of, say, a large gas turbine (over 35%) in typical driving conditions. It just can't be done in a car engine. The power/weight, acceleration, reliability/maintenance, and thermodynamic concerns are all working against it.
> and a more supportive infrastructure to help with repairs and maintenance
The model S's drive train has something like sixteen moving parts. And virtually no frictional contact except a few bearings. Compare with the hundreds (thousands?) of moving parts in a car engine, all furiously rubbing against each other.
Electric cars don't need maintenance, except for tires and brake pads. Motor failures are possible but very rare. The battery wears out gracefully and can be replaced, but doesn't really need it in practice. Under conditions of extreme stress the motor inverter might fail but even that is a freak scenario.
The kind of intensive and continuous maintenance you need with a gas car (oil changes, oil filters, air filters, antifreeze, transmissions, belts failing, radiators bursting, etc.) just don't exist in an electric car. It's just battery -> inverter -> motor -> gearbox -> wheel. An electric car is about as mechanically complicated as the engine _starter_ on a gas car.
You know nothing about physics and economics.
> If we're just going to shove the problem of fossil fuel burning off on someone else (i.e. a power plant), we're not actually solving any environmental,
Yes we are. A large, stationary power plant is always going to be way more energy efficient than a small car engine. A car engine is about 20% efficient; new power plant designs are capable of nearly 80% efficiency (especially when used for combined heat and power). 60% efficiency is typical for even older designs. If you crunch the numbers, driving an electric car powered solely by coal-based electricity puts out about half as much carbon as a gas car. This includes all the inefficiencies in power line transmission, batteries, etc. In reality this figure is actually much better than that since not all electricity comes from coal. But the point is that even if we were to use the worst CO2 emitter - coal - we'd still be making an improvement over gasoline cars, and it would be an economic improvement as well since the price of coal is far more stable than oil. Of course the goal is to do even better than that - to power cars using nuclear, solar, wind, and other clean energy sources.
> Right now electric cars are financially supported by the state, while gas cars pay for the roads
Instead of whining about tax on gasoline for road maintenance, car owners should be thankful they aren't being taxed for carbon emissions, which would be the case in a fair world.
Yes it's working perfectly, which is why everyone is talking about it.
Also you do realize that 'misdirection' of this sort is grossly illegal.
Every day my opinion of slashdot readers edges down a bit further.
> The Tesla Model S 85 kWh battery costs $44000.
Citation needed. Official cost estimates aren't known but Tesla probably pays around $20,000 for the battery on the P85D, and that's a big battery. http://insideevs.com/tesla-bat...
> My current car is 20 years old and requires very little in terms of operating expenses.
Cool anecdote bro. The actual data reveals that a 10 year old mid-price-range car typically requires $500/yr in maintenance costs; a 20 year old car requires $1000/yr on average.