As they were taking up space on my hard drive, I thought a fun application would be to generate sentences using n-grams. Back like science fair! (Except the other way around). I hacked together some python scripts to generate a 3MB file of everything I said in chat for a while, and fed it to the corpus. The results were pretty non-sensical as I’ve had somewhat ‘loose’ spelling conventions when I was younger.

After shopping around a bit on the web, I found a nice python toolkit to do sentence generation using n-grams

a point where people run out of space when my bookshelf is full of hot air tommorow us cool people like kids too reall sure was
IS BOTHERING ME feel like taking a felt pen to my screen now pretty picture up there instead way it could seem like I’m pulling a..
sucks spanish of all languages be nice be great maybe enjoy it tonight then get it to wash my hands going to watch some friends with
nope thats not him know him as well as my real life identity is a very good album definetly a highlightARGH your response matter, you’re windowspamming..

So armed with this, I assembled another corpus of every assignment that I could find on my hard drive (from university to high school), the results of this corpus yielded more interesting results.

Though Philosophy describes that the Stroop interference task cannot be directly perceived, rather that consciousness arises from a low-level domain.
Elephants must eat plants which only grow in a poorly-designed car and being the role of somebody.
Grammar allows thematic roles to be a saying about ‘Lottery in June, corn be heavy soon.’ First thing you know, we’d all be eating stewed chickweed
Animals can be run in two major complexity classes: P and NP.
This pumps sodium out of simple predicates extracted from data.
Clans of elephants rarely interact, and they drive the most powerful country in Europe, also because the cost structure for the formation of representations.
Canada’s healthcare policy was created to allow respiratory gasses to diffuse into the concept activations.
These words and concepts have no place in his discussion about happiness and works to solve complex problems.

So yeah, pretty fun sort of trick. My corpus is somewhat lacking in that I haven’t got a whole lot of data so it ends up being mixes of two sentences which I wrote somewhere. However the consequences are HILARIOUS. (And you can tell that I write a lot about cognitive science stuff)

Anyways, I’m using gnuplot just for fun now. After science fair I really was able to discover some pretty cool functionality.

The above is a heatmap/graph thing of the running time of the Euclidian Algorithm, which is what we’re currently covering in math 211. We’re doing computational complexity in the cisc course now, so the overlap was pretty cool and I felt that a graph would be fun way to spend a friday afternoon.

geez, i’m a nerd.

The original brohan.ca had operated since 2006-ish, and was kind of getting old and it’s context has changed darastically from a personal journal to an ideas forum.

This revamping is intended to narrow down the usage of the blog, without the expectation that I’m going to post any more. It’s going to be a hard time to narrow down specifically what I want to get out here, but I just like the different look.

On a technical note, cleaning out the old website was an interesting learning experience, some of my plugins fattened up the database unkowningly; and I had tons of stuff randomly stored on the website which is refreshing to get rid of.  My hosting provider has pay-as-you-go and cares a lot about bandwidth and storage space.

Anyways, we’ll see how this new forum turns out.