響きの声

This is something I recently (as in very recently) noticed, a disturbing trend on the quality of my study notes:

Basically, no notes for stuff you know extremely well, but the most detailed notes are usually for the stuff you know fairly well already anyway, and the stuff you don’t know usually has the worst notes. That’s not a very effective way of writing notes or an effective way of studying, since you don’t really learn anything anyway…?

So this morning the first years taking ENGGEN 131 handed in their solutions for a simplified knapsack problem. The story goes like this:

You have some number of boxes of a predetermined size, along with some number of weights, each with its own specific weight value. Your job is to find the best possible packing for the weights into the boxes so that you have a minimum number of unused weights.

I don’t have to take ENGGEN 131, but I decided to take a shot at it anyway.

After several miserable failures with ideas such as packing the boxes one at a time, I arrived at my semi brute force solution involving three (and a half) simple optimisations which can cut the number of combinations to test by quite a large number (eg. 4 billion combinations cut to around 300k).

Source: knapsack.c
Explanation: knapsack.txt

edit:

Also you can flip single line /* */ style comments to // comments while preserving whitespace at the front of the line using VS regex find and replace. Search

^{.*}/\*{.*}\*/.*$

and replace with

\1//\2

Task:
Write the code to sort data in ascending order in AVR assembly.

I thought it was pretty fun, but I only hit one snag. I kinda expected it to be harder :/

I would post the code, but then my fellow classmates would copy ;P

Summary:
Algorithm – selection sort
Instructions – 35 excluding the final end: rjmp end
Labels – 4

The best I can do with my code is 34 instructions, but I decided to split up one instruction to make things more clear.

It wasn’t exactly *hard*, but the lack of concise documentation meant I had to mess around for quite a while before I managed to get the UoA package running. So here’s the complete list of commands to execute:

$ apt-get install R-base
$ R
> install.packages(“s20x”)
> library(“s20x”)

edit:
I forgot that command line programs can’t display graphics, so all the plots and stuff won’t show up. So back to Windows install… >.>

http://up.ppy.sh/files/ds_closed.png
http://up.ppy.sh/files/ds_open.png

Still missing the sticker on the bottom, gonna do it tomorrow.

The probability of drawing the same name 3 times from the box is 0.0008. This event happened.

Question: Do you think the names in the box were well stirred?

My answer: I think the boxed was well mixed, but the drawer was cheating.

Result: -10

D: