Tuesday, February 9, 2010

School Life Survey 1 Wrap Up

School life survey 1:

Question: Are the claims of parents that social networking is harming today's generation true? How about the assumption that more "friends" really means more friends?

Hypotheses:
1. If more time is spent on social networking sites, then there will be more emphasis on social standing. (This hypothesis was just Null--no proof for or against since there was not enough data and I was attempting to use a Boolean-style approach, where Social Standing was 1 and Academic Standing 0. Came out like 2 straight lines.)
2. If more time is spent on social networking sites, then there will be less trustworthy friends.
3. If there are few friends on social networking sites, then there will be more hours spent on social networking sites.
4. If more time is spent on social networking sites, then there will be a lower weighted GPA.

These were hypotheses added on later (I made the above hypotheses on the spot as I posted my School Life Survey 1):
5. If there are more friends then there will be a lower GPA.
6. If there are more trustworthy friends then there will be less FB friends.

Data:

Hypothesis 1: Null.

Hypothesis 2:

Hypothesis 3:

Hypothesis 4:

Hypothesis 5:
(FB Friends to GPA)

(Trustworthy Friends vs. GPA)

Hypothesis 6:


Conclusion:
The data given suggests a strong negative correlation between the time spent on social networking sites and GPA. Also, there is a very slight negative correlation between the number of trustworthy friends/facebook friends and GPA. Otherwise, the data is too sparse to draw anything even semi-definitive.

Limitations/Uncertainties:
This data was taken largely from students from Mission San Jose High School in Fremont, CA. One or two points each come from Illinois and the East Coast (I do not know which points, obviously, as this was an anonymous survey, but my peers from those parts said they took the survey). Mission has a very high emphasis on academia, which leads to the relatively high amount of people who said they appreciated academic standing over social standing. Data is here: http://spreadsheets.google.com/pub?key=trtorCi2YDWnRugPcM3tkdg&single=true&gid=0&output=html

Wednesday, February 3, 2010

PEtS

Last week, Andrew Song and I expended a sizable amount of our beauty sleep time to design five web pages and write a competition paper all for Exploravision. I named it PEtS or Photosynthetic Ethanol Synthetes, and it was on how to make ethanol fuels viable by internal cellulosic degradation in the plant itself (which I had previously alluded to in my paper on Ethanol fuels (The Biofuel Fantasy).

The Idea in Development:

My idea was based on computer programmer logic :). I was thinking about how to make plants pretty much degrade themselves for the betterment of mankind, and I got into thinking about 'if' loops after I saw this graph that linked the proteins of Arabidopsis thaliana so cleanly with outer stimuli (it shows the signal cascade with light as stimulus).

Further Research & Finish up:

After my epiphany, I set to work to find a protein which could easily link to cellulose degradation at the end of the plant's life (because for obvious reasons the plant cannot degrade its cellulosic support while trying to develop into a plant). After one of the longest lapses in brain function I've ever experienced, I realized I could link the cellulose degradation to the Flowering Locus T Protein (which is transcribed and translated near the end of the plant's life when it will flower). Afterwards, it was a relatively simple idea: get the promoter code of the Flowering Locus T Protein using the corresponding sticky ends, insert before the cellulase degrading gene egl (from the bacterium inside a cow's rumen--this transgenetic modification of plants with egl has been done before in tobacco plants already as described by Kawazu, Sun, Shibatu, et al. in their article "Expression of a bacterial endoglucanase gene in tobacco increases digestibility of its cell wall fibers" for Journal of Bioscience and Bioengineering), then insert the modified gene into a "tamed" Agrobacterium tumefaciens with a specific antibiotic-resistant gene either to penicillin or kanmycin (Agrobacterium tumefaciens is commonly used to insert new genes in Arabidopsis thaliana), infect Arabidopsis thaliana protoplasts, and finally, select the modified protoplasts with the antibiotic the modified protoplasts have resistance to. When the Flowering Locus T Protein is transcribed, the egl gene is also transcribed (it has the Flowering Locus T Protein promoter right before it as well), so as the plant flowers, it degrades its own cellulosic material. This will theoretically obviate the cellulose degrading step in ethanol production, moving ethanol fuels one step further toward sustainability.

That was pretty much the bulk of the project. In the submission itself, I covered the technology thus far, the history of ethanol fuels, and what is necessary in order for PEtS to succeed, but otherwise, I summarized my idea quite thoroughly above.

NOTE: 2/26/2010 Andrew and I received an Honorable Mention for this project. I think it was impressive considering I spent a little less than 24 hours on the project (less than some of my school projects!). The title links to the Honorable Mentions list.

Monday, February 1, 2010

The Future Atlantes Part II

I recently posted "The Future Atlantes Part I" and promised a new installment. Well, here it is and it took a great deal of thinking and hair-pulling and teeth-gnashing:

The idea:

Make a 3D graph of the cities (x, y) with their population (z) shown as sticks scaled by 450,000 (the maximum population of the cities under 10 meter elevation was around 12,000,000). This way, it can easily show visual learners which parts of the world have the most people in danger of having to emigrate from their home place due to global warming (either with fair-warning or with a rough storm which inundates the whole place). This dilemma poses another question: how will all these displaced persons find a home or money (who wants a home stuck under the sea)?

My seemingly simple algorithm:

I'll start assuming you know what I did for Part I.
I made some extra rules of coordinates of cities to their corresponding populations. Then, I used the points of the cities under 10 meter elevation to be my x and y coordinates and the population to be my z coordinate (scaled down by 450,000). (This all had to be filtered as well against the not available data). Then, I graphed it with Graphics3D[].

A little more in-depth:

I created one module which took about 3 lines of code to convert a 2D point (the coordinates of a city) to a 3D line(which includes its scaled population and goes straight down: similar to the Filling option of ListPlot). That one module probably did the bulk of the work (well, with Map's help as well).

(Hopefully, I can show full functionality with a video on Wolfram Demonstrations or YouTube, as recommended by Dr. Choi).

The Graph without Labels and Scaled by 450,000:


Graph with Labels and Scaled by 100,000:


Future:

Scale the populations not to a stick length but to a colored point or maybe colored tubes.