October 2011
MagCloud is having a 25% off sale through tomorrow (the 31st). This means you can pick up my self-published magazine of fiction, poetry, photography, and miscellany, BLACK RABBIT, for just $5.40.
It’s a steal!
grngh, I am out money until the day after tomorrow
Nibbling on the tips of fingers.
It is not something I personally particularily enjoy, but hmm~
halcy replied to your post: Oct 26, 2011
“Leftist heaps”, heaps in which every element is equal!
halcy no
that is not how it works
Yes, sorry, of course! There is, however, regulation aimed at making sure every element is inserted into the heap at equal level!
In all seriousness, we never went over leftist heaps in any class - we went from binary straight to pairing heaps, then binomial and fibonacci heaps (Mostly for the heck of it, since for most practical applications, the added complexity of anything beyond binary heaps plus, recently, cache effects cancel all the runtime benefits for common data set sizes).
Well in leftist heaps, the tree is organized in such a way that things aren’t on equal levels. It is intentionally unbalanced so that the tree tends to lean to the left, which makes merging two heaps easier.
In our class we only did binary, binomial, and leftist heaps. Fibonacci heaps were briefly mentioned (i.e. “they exist”).
halcy replied to your post: Oct 27, 2011
Oh man… do you do mixed-signal processing at all at your school? I’m not sure if I should take a class in it, I am not really a hardware type, but it seems so interesting anyways…
I am not sure what that is; it sounds like it would be an elective or a later requirement for people doing the EE signals track, but I’m not really taking the signals track so I will probably not be taking it.
We mostly did pairing heaps because they are actually somewhat practical and because they lead well into externalized priority queues (priority queues which explicitly take a memory hierarchy - cache -> ram -> hdd etc - in account), and fibonacci heaps mostly as an example of a nontrivial runtime analysis.
Mixed signal processing is doing analog and digital signal processing in the same circuit (i.e. on the same die). There’s various things to take into account and compromises to make, apparently.
halcy replied to your post: Oct 26, 2011
“Leftist heaps”, heaps in which every element is equal!
halcy no
that is not how it works
Yes, sorry, of course! There is, however, regulation aimed at making sure every element is inserted into the heap at equal level!
In all seriousness, we never went over leftist heaps in any class - we went from binary straight to pairing heaps, then binomial and fibonacci heaps (Mostly for the heck of it, since for most practical applications, the added complexity of anything beyond binary heaps plus, recently, cache effects cancel all the runtime benefits for common data set sizes).
