terriko: (Default)
I got a really interesting query today that boiled down to, "How much math do you need to write code?"

The short answer to this is, "Not that much" or perhaps "it depends on what you want the code to do." But here's part of what I actually wrote back:


To be honest, the level of math required to write code is pretty small. A grade school understanding is often sufficient; there's a reason we can teach 7 year olds to program! Modern programming languages are much less math-oriented: I once spent an afternoon teaching my then 11 year old sister and her friends how to write dynamic database-driven websites, and the only math they used was to add up the scores on the "what animal are you most like?" quizzes they wanted to write.

The math in computer science comes a lot later: for deeper analysis of algorithms and running time, we use algebra and mathematical proofs in an academic setting. But... to tell the truth, relatively few programmers need or use this kind of deeper understanding in their day-to-day jobs. And in my experience teaching students, many people find this stuff easier to learn by doing, so they only really begin to grasp it *after* they have gotten comfortable writing programs.

In short: you probably have all the math skills you need to write code, and if you decide you want to do more hardcore CS later, it'll be easier to learn the math along the way anyhow!


There's some nuance there that I didn't really tease out -- the deeper understanding of algorithms and program behaviour is what characterizes the real "science" out of computer science. And maybe the world would be a better place if more programmers did actually use deeper analysis in their day-to-day jobs. But you don't have to be an academic-style computer scientist to write code! Still, it's a very interesting question, given that historically programming actually did require a lot more math, and our perceptions and stereotypes haven't really kept up with the reality of the field.

Perhaps it's time for me to write another presentation? ;)

(For context: my old slideshow about women, computing and math got included in this TechCrunch post about Racism and Meritocracy, so I've been getting a lot of mail, including the one that spawned this post.)
terriko: (Default)
My sister is the proud owner of the complete Star Trek: The Next Generation, and we've been re-watching first season while we eat dinner/do mending/make crafts/etc. It's much better than it has any right to be, actually, with surprisingly strong performances and decent writing. Apparently that first season cost a million per episode, and we're assuming a lot of that went into the special effects budget since they were done by ILM and they can't be cheap. (Although sadly, some of the effects are showing their age more than the actors' performances are...)

I can't remember which episode it was, but one of them contained some technobabble that I found fun because it was just a simple description of checksum bits within data. "That's not futuristic!" I joked, "We have those right now." I point vaguely in the direction of a library book. "Bet the barcode on that has a check bit."

Anyhow, in a case of "once something comes up, it starts coming up again and again" that I can't remember the term for (although there is one, and I think it's some german band name or something equally esoteric -- anyone know?) ... err. where was I? Oh, yeah. So, I just saw this graphic and it amused me because I actually think checksum bits are awesome because I'm that kind of mathematician *and* they'd just come up in the Star Trek technobabble (and I guess I'm also that kind of mathematician).

[Edit: Jay helpfully reminds me that the term I'm looking for is Baader-Meinhof. See Rob's comments on the gang it was named after, which was not a band. We should come up with a new name that's shorter, more etymologically related, and easier to remember!]

I've spent all week writing erudite paragraphs for my paper and I think I used up all my literary skill... okay, so I'm just tired and a bit lazy. But look! Math and codes!

Image via Mint.com
terriko: (Default)
Three links of interest from this week:

Conversations About the Internet #5: Anonymous Facebook Employee: What makes this story so entertaining isn't so much the content (which is pretty unsurprising, IMO) but the way in which it's presented. The drama! The intrigue! My personal favourite is describing eye-tracking, a fairly common technique used to analyze designs, as scary scary "psychological analysis." Seriously fun way of presenting what otherwise would be rather pedestrian information (OMG, Facebook keeps track of your relationships and behaviours! Like, oh, every other company that has any data about you...)

Programmers need to learn statistics or I will kill them all: You'd think there's no way that the essay could top the title, but it's actually a fantastic explanation of the problems many programmers have with statistics, as well as a reasonable rant about how little they care when they're told they're wrong. I've seen these mistakes in high-level peer-reviewed "scientific" papers in my field, and it kind of drives me (and many others) crazy. So if you're a computer scientist, go click that link and make sure you're not making those mistakes. You don't have to be stupider than slime mould, mathematically speaking.

An interesting side-note in that paper, for the women:

"Oh, and you wonder why I say, “he”? I never have this problem with female programmers. Maybe it’s because I’m tall (6’2”), or nicer to them, but they always speak rationally and are really keen to learn. If they disagree, they do so rationally and back up what they say. I think women are better programmers because they have less ego and are typically more interested in the gear rather than the pissing contest."

I leave interpretation of these remarks up to you. *grin* They don't have statistical significance anyhow. But either way, read the essay: it's a snarky but awesome and clear explanation of common statistical errors.

ProtectMarriage.com issues Cease and Desist for Prop 8 Trial Tracker logo depicting family of two mothers with two kids: ProtectMarriage.com threatens what seems to be a spurious lawsuit regarding a logo that is quite covered under parody laws. Prop8trialtracker.com hires the best lawyer ever, who responds with a rather impressive letter. I find it awesome that you can cite case history regarding entertaining stories like the slogan "Open up a Can of Woof-A**" -- I guess it's not entirely surprising that trademark case history will include a lot of funny/embarrassing examples, now that I think about it. Still, kudos to the lawyer who put together something so funny and clear on such short notice.


terriko: (Default)

February 2017

192021 22232425
26 2728    


RSS Atom

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Mar. 30th, 2017 06:36 pm
Powered by Dreamwidth Studios