Interesting followup to Serene's Trader Joe's post

Just found this article on the Freakanomics blog about the owners of Trader Joe's:

So, let me ask those of you who fall into categories 2 and 3: Who owns Trader Joe’s?

1. Some great California family full of surfers and gardeners.

2. A small band of communal farmers in Oregon.

3. A huge German discount-grocery chain best known in the U.S. for no-glamor stores often located in marginal neighborhoods.

Yeah, it’s No. 3. The company is called Aldi and, though I’d seen one or two of its stores in the past, I didn’t even know it was a grocery store. Then I read this very interesting Wall Street Journal piece about the company’s ambitious new plan for the U.S., which calls for 75 new stores this year. The article claims that Aldi is so good at selling cheap goods that WalMart couldn’t compete with it in Germany.

I did a little more investigating about Aldi. Apparently, Aldi specializes in running low-cost, discount stores - mainly selling house brands - in marginal neighborhoods in the US. But in Germany, according to Aldi's wikipedia page, :

95% of blue-collar workers, 88% of white-collar workers, 84% of public servants and 80% of self-employed Germans shop at Aldi.

That's some market share!!!

Would it be safe to say that Trader Joe's has cornered 80% of the YUPPIE market? A lot of people we know shop at Trader Joe's. Do you? Let us know, it would be interesting to find out ...

More Perl fun: Do the number of Sam's Clubs in a state correlate to its Obesity rates?

I wrote yet another perl script that scrapes data off walmart's webpage in order to see if there is a correlation between the number of Sam's Clubs in the state and the state's obesity rates. As you can see in the scatterplot below, there is... surprisingly, no correlation! :)

scatterplot:


Map visualization:

Who's your Perl daddy?

I am taking Lada Adamic's data munging class (officially titled SI601 - Data Manipulation and Analysis) this term and am learning lots. Check out what I managed to do this week - scrap our flickr contact list and display our contacts by the size of their contact list. Written in Perl, I am looking to extend the functionality of this script by looking at whether there's a correlation between the size of one's contact list (and hence audience) and the kinds of creative commons licenses they use. Hopefully with the extension of the script, I will be able to get data beyond just my contact list too :) Any suggestions or advice?

Oh yeah, from the many eyes visualization, it looks like 'Here is Katy' & 'Kevin Lim' are the big dogs on our flickr contact list. ;)