I took one of those online grammar tests

...so let's see if I can post the results here.

English Genius You scored 100% Beginner, 92% Intermediate, 100% Advanced, and 82% Expert!

You did so extremely well, even I can't find a word to describe your excellence! You have the uncommon intelligence necessary to understand things that most people don't. You have an extensive vocabulary, and you're not afraid to use it properly! Way to go! Thank you so much for taking my test. I hope you enjoyed it!

Automatic Gerrymandering

I generally look at gerrymandering as a problem that is to be avoided at all costs, but I thought I would have a little fun and come up with an "automatic gerrymandering" method that might yield bizarre representative districts but would still be relatively fair. Here's what I came up with:

1. Count the total votes cast in an election and use the Droop Quota to calculate the minimum number of voters to be represented by a single legislator.
2. Determine the winners of each precinct.
3. If a contiguous group of precincts have the same winner, and the number of voters in the precincts is equal to or greater than the cutoff, the winning candidate becomes a representative.
4. If a winning candidate has "surplus precincts" above the cutoff , the surplus becomes available for the closest candidate on down in each precinct as long as they do not eliminate an existing district by changing the winner, making the district non-contiguous, or dropping the number of voters below the Droop Quota. In other words, if a candidate won 100 precincts of equal size but only needed 90 to win, 10 precincts are available for redistribution to second-place candidates. Unallocated precincts then become available to the third-place candidates, and so on.
5. If a group of unallocated precincts is surrounded by one or more completed districts, then they are added to the neighboring completed district with the highest-scoring candidate as long as they do not alter his election. His surplus precincts then become available to other candidates (you are basically adding precincts on one side and making them available on the other).
6. If there are still representatives needing to be picked, unallocated precincts are combined into "super-precincts" until there is a winner. (I'm not sure what method of combination would be best, since there might be multiple ways of adding precincts together which would yield different representatives).

Of course, the downside to this would be the amount of calculation necessary to determine an election -- candidates who win outright are easy to pick, but choosing a winner among the other candidates is much more difficult -- but at least with this version of gerrymandering it's the voters in each precinct that makes the decision, not the politicians. You might also want to make a rule that unallocated precincts are first given to candidates who have won a sufficient number of non-contiguous precincts, and/or candidates with the highest total number of votes overall.

I'd be interested in seeing other ideas or refinements that others might come up with, To be honest I still prefer anti-gerrymandering districts (a vector quantization/Voronoi cell method would be great), but this method might yield interesting results in states like California and Texas. :)

Jewish holidays have such cool theme songs

(With apologies to The Knack and Weird Al)

Ooh my little pretty one, pretty one.
When you gonna celebrate Rosh Hashana?
Ooh you blow the shofar too, the shofar too
Blowin' in a year that's new, Rosh Hashana
Never gonna stop, eat it up.
Such a tasty snack. Honeyed apples are on this day
quite a bit like crack. My my my i yi woo. R R R Rosh Hashana...

A brief note about global warming

For those that take every sign of decreased rainfall as a sign of global warming, don't. Global warming might cause localized drought because of changes in weather patterns, but a higher average temperature tends to put more moisture in the air, just like raising the temperature on a covered pot on the stove. A disruption in the Atlantic Conveyor would cool temperatures in northern Europe which could definitely lead to drought there, but that will probably mean more extreme weather and flooding somewhere else -- you might see rain in the Sahara and force-5 hurricanes hitting the Atlantic seaboard.

That's the weird thing. Global warming is probably a bad thing for most of the places people like to live at present -- people like pleasant weather, and any disruption in the climate is generally going to be worse for them. For those now living in harsh conditions -- Siberia, Alaska, Northern Canada, the Sahara, Southwestern U.S., etc -- some places might actually have improved climate.

Space-filling Moore curves and data compression

Ever since I found out what a Huffman tree was, I've found data compression to be a fascinating subject. I've tried to apply the concepts to language ("Plan B" style -- for history of it, Google Lojban "plan b"), election methods (vector quantization of voting districts), and graphics (Google "hextar curve" or visit this Fractint L-system page).

Unfortunately, to quote the famous philosopher Barbie, "Math is hard." I'm fine with arithmetic, geometry, algebra, even basic calculus, but things like Abelian algebra, Hilbert space, and bra-ket notation make my hair fall out, and the math used in data compression can be a real pain without some kind soul to explain it.

Even so, I like the ideas involved. Another field I like is fractals -- I love things combining symmetry and uniqueness . For awhile, "fractal compression" held my interest, until I realized it was not any better (and often worse) than vector quantization. Wavelet compression is also quite interesting (once again, love the concepts, have a hard time with the math).

Then I saw some people doing interesting things with the Hilbert curve. Basically, a Hilbert curve is a way to map data in multiple dimensions onto a one dimensional string. It also tends to be compact -- points close together on the Hilbert curve are close together in space (though the converse is not always true). It is a plane-filling curve in two dimensions, and there is a three-dimensional space-filling version as well -- see Mathworld's Hilbert curve page.

Still, there is something about the Hilbert curve I do not like. It isn't closed for one thing -- the start and endpoint are on opposite sides of the square. There is however a variant known as the Moore curve where the start and endpoint are only one unit apart -- a single segment is sufficient to make it a closed curve. It can be seen here on this plane-filling curves page. I'm not certain if there is a 3-D variant of the Moore curve as there is for the Hilbert curve, but it does seem quite likely.

Which leads me to the compression idea. There are two major ways to attack the problem: top-down, taking the complete picture and subdividing with some function, then taking the subpictures and subdividing them, continuing to the level of individual pixels; and bottom-up, taking individual pixels and combining them, and continue the combination algorithm until you complete the picture. In addition the algorithm can choose for greatest difference, or nearest neighbor (I'll mention some variants later).

Top-down, greatest difference:
1. The Moore curve makes one loop of the entire picture, visiting each pixel once. If you imagine this as one big circle, with each pixel given a certain value, find the diameter that results in the greatest difference between combined values. For example, if the left side of a picture is black and the right side white, the greatest difference will be equivalent to a verticle line between them.
2. Take each subloop and join the endpoints, and find the diameter that results in the greatest difference in combined values. Repeat until you have subdivided the picture into pairs of pixels.
3. Run your compression algorithm on the final data sets (more on that later).
4. If you are interested in video compression, you can like of a video stream as a three dimensional block. A 3-D version of the Moore curve can be used, with the same subdividing and looping.

Bottom-up, least difference:
1. Take the Moore curve of the image or video, and compare the absolute difference in pixel pairs and choose the smallest value. For example, if comparing the pixels A, B, C, and D, you would compare |A-B|+|C-D| and |A-D|+|B-C| and choose the smaller of the two.
2. Combine pairs of pixels the same way, and choose the combination with the smallest total.
3. Continue until you have two halves of the image or video, and run your compression algorithm on the data sets.

Either method should yield good results. Basically, the bigger the blocks the more we want there to be a difference between them (because like colors will preferentially be in each block), while the smaller the block the more we want them to be the same (to improve data compression). Which brings us to the final method.

Brute-Force, maximal compression:
For a picture 1024 x 1024 pixels, you are looking at a little more than one million ways of partitioning the image. For a 1024 x 1024 x 7200 seconds x 30 FPS video, there are roughly a quarter of a trillion ways to partition the image. Short of some kind of quantum computing or dramatically faster processors, a brute-force method of video probably won't happen.
On the other hand, a test of different partitioning algorithms over a simpler subset would probably be a good idea. A 1024 x 1024 x 1024 cube -- equal to a 1024 x 1024 x 34 second movie -- can be partitioned with a Moore curve with roughly 1 billion possible partitioning methods (think a 30-bit number). If one were to calculate all the various possibilities as well as how much they compressed, and then compared them to the methods above, one could get a feel for the most efficient algorithm.

I'll go over this a bit more when the inspiration strikes (grin).

A rant about the legal system

I recently wrote the following rant in response to the following statement:

jlowejd
Not for nothing, but some of you farkers need to STFU. You have to understand that the 95% of us who are just plain lawyers absolutely freaking HATE that other 5% that make the rest of us look like shiat in the eyes of the uneducated public.

Is that like the 95% of Muslims who aren't terrorists?

/Some goats are trying to cross my bridge. BRB.

Okay, I'm being facetious, but if 5% of the doctors started spreading vials of contagious disease around in order to drum up business, I would hope the other 95% would do more than feel offended. Where are the groups of lawyers dedicated to simplifying the legal system so the average person can understand it? Which ones are trying to close the loopholes that unscrupulous lawyers use to extort money from others? I'm sure there are some -- though nowhere near 95% -- but apparently they have a pretty poor publicity department. You don't see physicists trying to complicate physical laws just for the hell of it -- they gain fame (not to mention Nobel Prizes) if they can explain things better.

Unlike some, I actually see a need for lawyers -- a day in court is better than a duel at sunrise. I just wished they were more concerned with clearly and concisely defining person's rights and responsibilities in society rather than trying to increase the complexity (and the number of billable hours) of the legal system.

As a thought, why not require law students to write a thesis taking a complex existing law and simplifying it so the average person can understand it? You could add points if it removes loopholes, and an automatic passing grade if the law is submitted to the legislature and approved.

/I'm well aware my rant will attract a bit of heat, but when it takes longer to read the shrinkwrap agreement on a program than it does to install it, something is wrong.

From Income Tax to National Sales Tax

I live in Oregon, and for the past three decades or so I've been told by the politicians in power how wonderful a sales tax would be. Yep, I live in one of the few states that do not have a sales tax.

Well, if a state sales tax is so wonderful -- and believe me, Democrats have tried to push is at least as hard as Republicans (remember, this is Oregon), why would it suddenly become regressive when applied nationally?

My only problem -- and the reason I vote with majority of other Oregonians to defeat the sales tax every time -- is that politicians want it in addition to the current taxes, not instead of them. Most of us learned our lesson after we voted for a lottery -- the money was supposed to go to education but instead a huge chunk found its way into the general fund.

As for me, there are some good points to a national sales tax -- tax shelters for the rich are a bit less effective when their purchases are taxed (and they do like to buy things -- offshore investments don't do them any good until they actually spend the money), plus those in the "hidden economy" (illegal aliens, drug dealers, whatever) and those who cheat on their taxes still like to buy things. As long as all transactions are taxed (including income, which represents a sale of service) so there are no loopholes, I'm willing to at least consider the idea.

Healthcare reform

There is actually a congresscritter who wants to make the following amendment to the constitution:

"all persons shall enjoy the right to health care of equal high quality."

U.S. Rep. Pete Stark, D-Fremont

As others have pointed out, there are other even more basic requirements than health care. Why not an amendment:

"all persons shall enjoy the right to food, clothing, and shelter of equal high quality."

What about legal services? The legal system is one of the functions of government, but while poor people charged with a crime can get a court-appointed lawyer, I'm pretty sure most of them aren't the Johnny Cochran's of their class. Shouldn't we have an amendment mandating legal representation of equal high quality? We can put price caps on it and everything, and if we mess it up and lose our brightest lawyers at least is wouldn't be as bad as losing our brightest doctors.

All this may seem like I'm against universal health care. Actually, I wouldn't have an objection to it as long as the following criteria are met:

1. Preventative services (vaccinations, prenatal care, etc) are funded before anything else.
2. Medical expenses for things caused by lifestyle choices (obesity, drunk driving, smoking) are paid by the patient.
3. Medical services are preferentially given to those with a greater chance of contributing to society -- young people get organ transplants before old people, with criminals on death row at the back of the line.
4. Some mechanism for restricting "vanity" and unnecessary care (facelifts and hypochondria are kept out).
5. People who use more health care pay a greater percentage of their take-home pay for insurance.

There are probably other criteria as well, but that would be sufficient to start with.

Taking the long view

Sometimes I want to throw up my hands in disgust at people. The "long view" in Congress is limited to the next election, every problem is seen through a partisan lens, and we just go from crisis to crisis without planning for things ahead of time.

You know what would be great? If Congress would make a list of all foreseeable threats, their probable distance in the future, the best long-term solution, and how much money would be needed per year to fix the problem. Just picking numbers at random, you would have something like the following:

Universal Heat Death/Big Chill (total destruction) 100 trillion years -- unknown, $10 million/year basic research
Sun going Nova -- 5 billion years -- leave solar system -- $10 million/year space science
Earth becomes uninhabitable -- 500 million years -- leave earth/$50 million per year space science
Earth is struck by an asteroid -- 1 million years -- $25 million/year global watch to determine threats and $25 million/year for space science research
Storage of nuclear waste -- 100,000 years, $100 million/year storage, $100 million/year fusion research.
Global warming (climate change, coastal flooding) -- 100 years/reduce greenhouse emission/$1 billion per year industry loss/$100 million per year research
Social Security bankrupt -- 50 years, {very large number} billion to make it perpetually solvent.
National Debt/Depression / starting at 10 years/ Balanced budget in place before then, pay down the debt over next 50 years

/etc

(Yes, I just pulled numbers out of thin air. I just wanted to give an example of what I mean by long-term planning)

As new information becomes available, simply adjust the budget accordingly -- if we discover cold fusion, there's less need to find out a cheap method of disposing nuclear waste. Sure, that means that my generation might have to pay for things that will never directly benefit us, but I know the country would be a lot better off today if the founding fathers had had the foresight to fix certain things from the start -- a "balanced budget" amendment to the Constitution, perhaps, prohibition of slavery, etc. -- so I'm willing to make that sacrifice for future generations.

Education in America -- year-round schooling

One easy improvement to our current school year is to get rid of summer vacation and have more frequent but shorter breaks throughout the year. Some schools already do this -- sometimes called year-round schooling or 45/15 (for the number of schooldays on and off, or 9 weeks of school followed by a 3-week vacation).

Such schooling has been around for quite a while -- I attended a 45/15 class from second to third grade at Highland Elementary in Norco, California in 1972-73, and still remember it fondly -- and in my opinion is far superior to the current method used by most schools.

It can be improved, however. If we are going to keep roughly the same number of school days, why not move to a 4-day school week (Monday-Thursday)? Twelve 4-day weeks equal 48 days per quarter, with a week off between quarters. The occasional Monday or mid-week holiday (Thanksgiving, New Years) would bring it down to 45 days. Make certain one break covers Christmas and the rest of the breaks will fall into place, one per season -- you'd have Spring, Summer, and Fall Break as well as a break during the holidays.

The nice thing about this is that it keeps Friday free -- which is not only where many public holidays land, but the day many schools set aside for sports competitions (football and others). This way, sports and extracurricular activities are less likely to interfere with schoolwork.

As a side note, the truly optimal method of year-round schooling IMO would be to have a 3/2 schedule -- three days on, two days off -- with appropriate vacations during the year. Unfortunately, that is extremely unlikely, since we'd need a 5-day week ( a calendar I'll blog about at a future time mainly as a fun idea). With a five-day week, you could have 18 weeks (16*3 or 48 days) per quarter, with two weeks off between quarters and an extra week off between school years. Since education would be less compressed, with plenty of breaks, students would have a better chance of learning the material required.

The cost of war in Iraq

I've seen articles on the value of human life from an actuarial and statistical point of view. Most people value their lives at $7-$10 million. The U.S. uses about 18 million barrels of oil a day. If you are able to decrease the price of oil by $1/barrel, it's worth (in statistical terms) about two lives per day -- less, if you figure the cost of transportation, training, and equipment.

Now, if you figure $200 billion dollars and 1500 lives (taking the higher $10 million price tag per life), this works out to the equivalent of $215 billion (or conversely, the equivalent of 21,500 lives) over two years. If we take current oil prices of $53/barrel (total around $700 billion over two years), then we should expect oil prices to be about 30% lower than they would be otherwise. Somehow, I don't think they are.