Monthly Archives: August 2012


I like data. I believe in collecting data and analyzing it to see interesting trends. And it’s always sad to lose data, because then I lose opportunities to analyze said data.

One piece of data that I find interesting is sleep. The amount of sleep that I get each night strongly affects how happy and productive I am, and I want to analyze this data to see if there is anything that I can learn about my sleep patterns. So I’ve been recording my nightly sleep schedule — when I go to sleep and when I wake up. The hope is that I can eventually produce pretty graphs of when I sleep. It pains me to think that some people could be recording their sleep schedules too, but instead fail to do so out of laziness.

I’ve been recording sleep data since 13 March 2012 (excluding naps, e.g. boring classes; all times approximate and rounded to the nearest five minutes), and I can already see some patterns. The full analysis hasn’t happened yet (I plan to wait until I have more data), but here’s some preliminary statistics (as of today, August 26).

Average since March 13: 7:59:21
Maximum: 12:29 on March 29
Minimum: 0:00 on March 24 (thanks to red-eye flight)
Minimum positive time: 2:45 on June 3 (thanks to ARML)
Standard Deviation: 1:27:08

I think I’ve been sleeping more (and more normally!) over the summer:
7-Day Average: 8:41:25
15-Day Average: 8:39:20
30-Day Average: 8:29:52

Let’s hope that this trend continues!

There have been a couple of erratic weeks:
Maximal 7-Day Standard Deviation: 4:14:15 (week ending March 30)
And some very regular weeks:
Minimal 7-Day Stanford Deviation: 0:04:30 (week ending April 15, a week of times between 8:00:00 and 8:10:00)

Conveniently, recording sleep data also makes me realize when I’m not sleeping enough and motivates me to sleep more. Which is a good thing.

Two New Computers

Thanks to Matt for making me feel guilty about failing at blogging. Lesson: guilt is far more effective than to do lists.

I got two computers in the mail this week!

The first computer is an old Lenovo Thinkcentre that was acquired from eBay. It’s going to be the Stanford University Mathematical Organization‘s new server. We’ll use it to do a whole bunch of things, including but not limited to a number of jobs for the Stanford Math Tournament (registration, grading, and tons of data analysis). I installed Debian Squeeze last night, and it should be fully functioning within the next few days.

The second computer is a Raspberry Pi! I ordered a Raspberry Pi from RS on June 29, and after increasing amounts of frustration, cancelled that order and ordered again from Newark on July 20. After another few weeks of waiting, it’s finally here! It looks really pretty. Unfortunately, I’m still lacking various cables, an SD card, and free time, so I can’t play with it yet. With everything else that I should do this week, that’s probably a good thing. I’m still trying to decide what to do with my Raspberry Pi. Any cool ideas? Let me know!

Before I got my Raspberry Pi, I had heard that it was credit-card-sized, but I was still surprised to see how small it is. It’s amazing that something that small could have so much computing power — this computer is more powerful than the best supercomputers from only a few decades ago.

Here’s a picture of the two computers:

SUMO server and Raspberry Pi