tungwaiyip.info

home

about me

links

Blog

< February 2014 >
SuMoTuWeThFrSa
       1
2 3 4 5 6 7 8
9101112131415
16171819202122
232425262728 

past articles »

Click for San Francisco, California Forecast

San Francisco, USA

 

Paul Graham on Cities and Ambition

I was rereading Paul Graham's article on Cities and Ambition. His idea is each city project a specific message to the people. The message attracts and motivates people to a specific kind of enterprise. New York is about making more money. Boston is about becoming smarter. Silicon Valley is about to become more powerful (by means of innovation). Paris was a about intellectual but today it is more about arts. This is fairly consistent with Richard Florida's thesis that location has great influence on what people do and what they could achieve.

One astute observation that Graham put forward is you don't have to live in a great city the whole life to benefit from it. Where you grew up and went to school is not so important. Instead, what matters is a critical period from the early and middle career. The example he gave is the Impressionists painters. They were born all over France and died all over France, but what defined them were the years they spent together in Paris.

I come to Silicon Valley at the age of 27, still in my early career. This has happened opportunistically. When I was younger, I have never imagined I would emigrate to a distance place. This location has indeed redefined myself. Now I am working in the forefront of technology in the cosmopolitan where innovation happens daily. Had I stay in Hong Kong past 30, the window will probably close. I would not have envisioned doing things I am doing today. Biologically I am the same person with the same capability. But I would not have the same ambition compare to the peers in the Silicon Valley.

2014.02.27 comments

 

Mathematical Intuitions

During my online course study, sometimes the instructor would say, intuitively, the reason go such as such. We spend a lot of time doing proofs that are the primary insight and knowledge. But the instructors also speak of developing an intuition on the subject.

It dawn on me that these intuitions are the superpower that separate people who "just know" from the rest. This is not some short cut but the critical skill of mastery. Some really simple algebraic intuition are x2 grow faster than x; 1/x decrease as x increase, etc. This is obvious to those who know them. But to others, they may not see this immediately and may need to think or make calculations. These students don't have the intuition. They may have figured out at the end, but they do not have the mastery.

I was fairly good at math at school. Now I am learning a lot about probabilities, a somewhat a new territory to me. A lot times my intuitions plainly fail. I stare at some simple symbols or equations and cannot make out what they are. I may have to slowly go thought definitions and try to work them out. But after working on these topics for many months, I begin to be able to see more things intuitively. Mastery in mathematics is not just knowledge in formal proofs. A lot of times it is to develop the intuition so that you just know.

2014.02.25 comments

 

Make Useful Charts

Having been with Bay Area Bike Share for a few month, I am glad with their green bikes available around downtown. I also appreciate Bay Area Bike Share's openness in sharing their system metrics. However the system chart on their web page do not provide much useful information.

Official Chart

The slowly raise bar chart shows cumulative trips taken since launch. The chart may look boring, but as least the upward trend looks comforting, right? Wrong! A cumulative chart by construction can only go up. The comfort feeling is misleading. A more useful way is to chart the number of trips taken by week. Also, data from 5 cities are available individually. So why not plot them side by side for comparison. I did a little work to make an improved chart below.

Improved Chart

Some information immediate obvious from the chart is that nearly all trips were made in San Francisco, despite it only has half of the resources deployed. Clearly the bike share program is not working out in the Peninsula at all given their negligible usage. Secondly, there are ups and downs in usage that cannot be observed from the upward looking bar chart. After the first few months the usage has peaked at about 6,000 to 7,000 trips per week, then fallen sharply in December due to holidays. The January number has yet to recover to reach the usage last year.

This is not mean to be critical. It is just to demonstrate how an useful chart can inform us.

This is the link to the simple source code by Pandas.

2014.02.24 [] - comments

 

Why I Hate Whatsapp

Whatsapp's was sold for a jaw-dropping price, with more 0s than any normal people can count. I have no idea if this price is reasonable. As a user, I just want to talk about why I hate Whatsapp, as well as the entire class of mobile only proprietary messaging apps.

One immediate drawback for mobile messaging apps is that it is really hard to type on smartphone! Say all you can about how an on screen keyboard is as good as or better than physical keyboard. My experience is nothing like that. Correcting typo, which is happening all the time, is atrocious. It is very difficult to position the cursor back and fro precisely. Often it is better off to delete the whole word to correct a single mistake. Many times I simply give up and let readable typo, capitalization or punctuation error happen. My mobile efficiency is fairly low. On average I am able to compose about one to two messages per minute on a cellphone. Composing message longer than one sentence start to get awkward. For many people who carry a smartphone all the time but not necessary have access to a computer, the convenience outweigh the issues. But I am essentially in front of a computer all day. The inadequacy of mobile keyboard is obvious and really frustrating.

Another main reason I dislike Whatsapp is that it is a proprietary walled garden. I fundamentally distrust proprietary messaging network. If this were an open network running on standard protocol, someone would have create a desktop or webapp alternative to allow more access and fix the keyboard problem. Whatsapp decided they are a mobile only application. So we are stuck with whatever they made. They can decide on what they charge. And if the company fold one day or decided they do not want to provide the service anymore, you network will be gone with them.

The idea that Whatsapp will disappear in the short term seems ridiculous given its astronomical sales price. But I do not think its eventual decline will be far fetched. I have gone through a long list of once high flying instant messaging apps like ICQ, AOL, MSN, Yahoo messenger etc. People abandon them eventually. I have only recently disconnect from Yahoo Messenger. I left only after most of the people I used to chat with have long gone. By this time, majority of messages I receive from Yahoo Messenger are spam and porn. This is actually not an unusual end for these commercial networks.

For now I am an unwilling user, because many people, in particular of those in Hong Kong, use Whatsapp heavily. I bet with high confidence that within the next 10 years, some other app will come along and displace Whatspp. And I hope that this will be a message system base on open network.

2014.02.21 comments

 

Honorary Retirement - Collins Dictionary Plus

I have removed the my Collins Concise Dictionary Plus from my top drawer to put it away to some other book shelves. My drawer has very little space. It is sort of prime real estate for my stuff. Since I have not really used the dictionary in the last 10 years, it does not make sense to for it take up valuable spaces.

Collins Dictionary Plus

I bought the dictionary in 1992. For me it was quite a find back then. The "Plus" in the title refers to the fact that it has extensive encyclopedic entries for biographical, scientific, political, geographical, historical, and the arts. Basically it was my Wikipedia at the time. 1992 is around the time I started to read a lot of English materials for general interests, like books and a whole stack of National Geographic. My English comprehension was quite limited. I pretty much have to look up a word for every paragraph I read, if not for every sentence. And I really appreciate the brief encyclopedic entries that give some background on prominent people, places or concepts. You know, it is like a mini-Wikipedia.

At the beginning, the dictionary was always by my side. Slowly I use it less and less. My vocabulary was improving. Besides, it was rather cumbersome to lookup dictionary all the time. I rather keep on reading and not to interrupt the flow. Later the web has come. If I have to look anything up, I go to the computer and I can find load of information instantaneously.

So the dictionary has settled in the drawer inactively. It is time for some spring cleaning. And it is a time to reminisce how wonderful it was to me.

2014.02.17 comments

 

Tech Buses Related to Violent Crime

There is a news story today about the S.F.'s 'clusters of affluence'. The headline says "Where divisive shuttles roam, affluence appears". The study come from a mapping of known tech shuttle stops overlay with map of restaurants and cafes. It shows that there is a pattern of overlapping between the two. And so here comes an Sunday article saying how tech bus form affluence clusters and fuel gentrification.

Not satisfied with the conclusion, I decided to make more maps. I have used the same tech bus stop listing, but instead I overlay it with incidents of violent crimes, murders, aggravated assaults, etc. The visualization is shown below. Lo and behold, there is a correlation between tech buses and violent crime!

Lest the bus protesters going to jump on this and accuse Google buses lead to violent crime, I would note that I have picked the crime data from 2008 and overlay it with the bus stop data from 2012. So the causality does not work. Instead, maybe someone would want to explore headline like "Fear of escalating murder incidents, Googlers seek refuge in private shuttle buses".

Chris Walker, the maker of the "Cluster of affluence" map, uses the concentration of restaurants as an measure of affluence. I do not find this metric as useful. Notice there are a big blob of restaurants in Chinatown and Tenderloin. They are clearly not affluent neighborhood. The more interesting case is in Mission. Is it affluent? On the one hand, the rent is rising and restaurants and cafes are flourishing. On the other hand, the demographic is still largely low income. Where fancy restaurants opens, you will find just as many fast food joints. I hesitate to label Mission as "affluent". Instead I think the more appropriate label should be mixed.

At the end, both of these maps are actually mapping of high density urban area. If you look closer at the restaurant map, they are just the map of well established commercial corridors that preceded Google. They are not created by the tech buses. These are places with high concentration of people and activities. So there are more cafes, more restaurants, and more bus stops. But these are also more homeless, more drug dealers, and more violent crime. They all happen in the same dense urban core.

The true story, I contend, is the revival of urbanity. Unlike the last generation who prefer spacious and secluded home, private cars to work and dine, this generation come back to the walkable urban core. To the surprise of the last generation, they eschew driving and content with commuting by bus. They also confront the crime that have driven people away. The shuttle bus stops simply follow where people go.

In my opinion the phenomenon of revival of urbanity is decidedly positive. People are choosing more sustainable live style. They reengage with the community rather than seeking isolation. They reversed the trend of segregation and mixing with other demographics again. While these changes inevitably causes frictions and challenges, let's not forget the social problems of the yesteryear, of white flight to suburbs, leaving behind the urban core to poverty, crime, and decay.

Links

[1] Dotspotting - "Foursquare Shuttle Stops", a sheet of dots by zach
http://dotspotting.org/u/939/sheets/2227/#c=12.00/37.7650/-122.4235

[2] sanfrancisco.crimespotting.org crime data API
http://sanfrancisco.crimespotting.org/crime-data?format=csv&type=AA,Mu,Ro,SA&dstart=2008-07-01&dend=2008-08-31&count=5000&bbox=-122.518330,37.708062,-122.373447,37.812560

[3] Visualization
http://cdb.io/1fPOvkR

2014.02.09 comments

 

MOOC 2013 review

I have been a devotee of Coursera since it has launched in 2012. 2013 is the second year I am really engaged in online learning. I've completed 13 courses in one year alone. It is time to look back to the year.

The year start with Calculus: Single Variable from Robert Ghrist of UPenn. I have revisited many calculus topics, some I have learned in the past and some are new. It was so long ago when I have last done it. I am very delighted to reconnect with them. Of course calculus is required in many other courses, so this has great practical importance also. Ghrist's lucid presentation, his enthusiasm in mathematics, and the presentation full of animations really brings the math alive. This is an exemplary online course, a hard act to follow. I am eagerly awaiting for Ghrist's follow up course on multivariable calculus!

The other accomplishment is I have learned to write mathematical formula using Latex. Being able to author math formula is an empowering skill, much like when I have learned to input Chinese a while back. It is a quintessential literacy skill in the digital environment.

I like to mix technical classes with humanity class for variety. In this case it is The Modern World: Global History since 1760 from Philip Zelikow of University of Virginia. Rather than recounting historical events that leads to the modern world, Zelikow put us in the shoes of the people in the era, showing us the challenges they are facing in the time and their responses to those challenges. It gives me new perspective for these societies in many ways. This is a highly enjoyable 14 weeks journey.

Coding the Matrix from Philip Klein of Brown University teaches linear algebra by programming. I have previous studied linear algebra from textbook from a more mathematical perspective. Klein show us the many practical applications and how we can program them. This is an exciting skill to learn. This eight week course is rather short though and does not cover many important topics.

Hong Kong's institutions have also begin offering MOOC classes. I have enrolled in James Lee of HKUST's A New History for a New China. His approach eschews the chronological narrative for an analytic approach, using historical statistics to understand social aspect like wealth distribution in the past. It may sound dry but I find it very refreshing. One accomplishment is that I have initiated a translation project in the class. There is a class reading, an extract of a book's manuscript that's only available in Chinese. I invited volunteers to translate it into English for the larger audience. With about 10 contributors, we manage to translate much of it in 2 weeks, just in time for other students to learn.

One new strand of study I have started in 2013 is bioinformatics and genetics. I was rather fascinated with microbiology and how it can be approached as a computation problem. I finally got my start, first by working through the algorithm problems from the rosalind.info website. The Rosalind.info team has subsequently developed the UCSD MOOC course Bioinformatics Algorithms. As the same time I also enrolled in University of British Columbia's Useful Genetic to gain a biological perspective. My background is much stronger in algorithm than genetics. Still I get an appreciation in biology and learned that there are much more about life science yet to be discovered.

Other honorable mention are Data Analysis from Jeff Leek of Johns Hopkins University and Functional Programming Principles in Scala by Martin Odersky of École Polytechnique Fédérale de Lausanne. It was a great year. I highly appreciate all the wonderful courses they put together and offered to the world for free.

2014.02.02 comments

 

Made with PyBlosxom