How to Choose a Mac

As I've mentioned in previous posts about laptop battery life, my 15" MacBook Pro (MBP) is over 3 years old. I'm now starting to seriously look at buying my next computer. How do you choose which to buy?! A few things conspire to make this a stressful decision for me:

1) I'm very particular about buying things like computers.

2) I rely on my portable computer as my only machine (by choice/design) and use it day and night for both astronomy and my side projects (like my blog!).

3) I have been traveling a good amount, around 3 or 4 conferences and a few weddings per year. This is going to continue (or maybe increase).

All of my past Apple machines have been excellent, and I've selected them each with care and forethought. Usually I wait to buy until the machine that makes sense gets an upgrade, and there is a whole niche industry in predicting/discussing such timescales. For example, I purchased this MBP shortly after an update, and right before I was starting my PhD program. I knew the best value for a high performance portable computer would be the 15" MBP model.

This time I'm very torn. I had been hoping that Apple would release a 15" MacBook Air, essentially unifying their laptop line. Largely this is what they've done, with the new slimed down and beefed up MacBook Pro Retina (MBPr).

The MBA has also grown, for its part. What used to be a big netbook has now blossomed into a respectable workhorse. For things like writing, traveling, and sheer beautify, the MBA is unmatched in my eyes. With a few affordable upgrades, the 13" MBA is on my radar as a real contender.

Both the Retina and the Air are capable of handling my day-to-day workload, both have been recently updated, both argue for portability. How to choose? MacBook Pro Retina or MacBook Air? Go for the top-of-the-line expensive heavier animal, or the urban and sophisticated marvel? I've asked people for their opinions, but the feedback was limited.

I decided to see if looking at numbers could help the matter! Is there an objectively better choice? Is one Apple laptop truly more desirable?
Resolution and RAM to Weight ratios as a function
of price for various Apple portable computers.
Here I've shown the screen resolution to weight ratio, and RAM to weight ratio, both as a function of price. These seemed to me the most important characteristics of a new laptop. As expected, the MBPr completely overpowers the resolution argument, while the MBA with upgraded RAM and CPU dominates the memory to weight ratio (with the MBPr in a respectable 2nd place).

These were interesting, as were other comparisons of the stats, but a clear winner didn't emerge. I needed a way to combine all of these characteristics, resulting in an "Awesomeness Parameter" that I could use. Then, simply pick the machine with the highest Awesomeness: easy!

I toiled for hours minutes, weighting the 4 parameters in various ways, and finally arrived at this:

The "Awesomeness Parameter", which includes Price, Weight,
Pixel Count, and RAM, shown as a function of the price.

The parameter "Q2" includes of all these factors: price, weight, resolution, and RAM. Each quantity is combined using my proprietary algorithm. The parameter is normalized to my current MBP (red), and thus Q2 = 1. The red dashed line shows how Qchanges if you simply decreased the price of my current machine.

While this statistic is meant in utter jocularity, it (strangely) accurately describes my feelings about the current laptop line. I've been very impressed with the 11" MBA, for instance, though it lacks the CPU power and HD space I need. Naturally the future Q3 parameter will include these stats as well.

The two top contenders are VERY close, with the MBPr scoring a 6.5 and the upgraded MBA a 6.7. By this metric, I should clearly purchase the upgraded 13" MBA...

Of course, I'm not guaranteeing my proprietary algorithm completely solves the problem. Still, it presents a curious idea: can I accurately quantify/model my desire for a computer, wrapping the decision up in to a single number?

What do you think? Which computer is better? Can we believe in a quantity like Q2? Chime in!


I made this animated gif on the plane ride back from a conference last year. Someone had it in a talk I saw and I couldn't find a version of it online, so I had to make my own. I need to re-make it with proper resolution, as it looks grainy and ugly. Still, it nicely demonstrates how parallax works in astronomy. You know the stars that move left/right the most are close. The stationary red circles are my "very distant background quasars". In the bottom left corner you see a representation of the Earth (blue) going around the Sun (yellow).

UPDATE: it's 2018 now (holy smokes, almost 6 years later...) and I re-discovered this GIF on my computer. Bonus: sometime in 2014 I actually did re-render it with better resolution. Also the code is now public. Enjoy!

Plots that Changed the World

Be sure to subscribe for updates on this and all my other data analysis projects! 

I spend a lot of my time looking at plots (aka graphs, charts, figures...)  If done carefully, graphs can be informative, artistic, personal, edifying, and even clever/witty. So often they are hasty, convoluted, gritty, and confusing. Frequently you encounter figures that seem to have been made out of spite for the data, or to merely satisfy tradition in the field. 

There are a wealth of opinions on what constitutes a good graph, and the philosophies behind creating them. Since there is commonly little formal training for scientists in the art of visual communication à la graphs, I think there is much still be said and written on the subject. I hope to bring some outside expertise and opinions to If We Assume on this very matter. 

One aspect about graphs that has entranced me is their ability to tell a story. A universal tenet of effective figures is that they must tell the story better than any other medium, better than a text description, a table, or an equation. Otherwise they are redundant. By visualizing the story (actually the data) we make it more human, more digestible. It touches chords, associates with memories, and should affect us like a good parlor trick.

The title of this post is a double entendre. I am presenting five plots/graphs that are famous, and have changed the world in a way. Each figure also represents a plot/story, capturing and quantifying the human saga throughout some piece of history. Most are not attractive enough to merit studying their aesthetics unfortunately.

Stat Trek

No comments:
USS Enterprise (refit) - from Wikipedia
It's a well known fact that I'm a big Star Trek fan, having spent way too many hours in childhood (and college) watching every episode and movie. Between Trek and my enjoyment of statistics/visualization, I guess you could call me a Data Nerd... oh man, I kill me.

Thus, it's only natural I should combine these passions, and bring to you STAT TREK