tinselt0wn posted:

im "learning" j again

quicksort=: (($:@(<#[) , (=#[) , $:@(>#[)) ({~ ?@#)) ^: (1<#)

suck it legibility havers

MononcQc posted:

While we're in algorithm talk, I'd like to ask if anyone has a few tips here.

I'm working on doing some advertisement targeting at the moment and I've been working on implementing Interval IDs to evaluate Boolean Expression Trees (PDF) to allow me to do some nifty efficient arbitrary matches, getting close (~3x) to compiled-code speed when interpreted on a few reasonable queries.

The problem I foresee is that it's still gonna be lovely expensive to run this stuff over hundreds of thousands of candidates. I was wondering if anyone working in different areas had similar problems of doing complex matching (hard to index) over medium-sized to large data-sets with ideally sublinear complexity, and if so, if you'd have any pointers .

It seems to me pub/sub problems are somewhat similar, as well as match-making for dating sites or online games and whatnot. I'm sadly under low-latency constraints (under 20-30ms if possible), but any source of inspiration is welcome.

ahhh spiders posted:

in what context

ahhh spiders posted:

not in lua, because you wouldn't do those retarded C things in lua

rotor posted:

how about you shut the gently caress up

Sweevo posted:

shaggar, like a stopped clock is right twice a day

tbc has a php script which moves the hands whenever the clock gets too close to being correct

tef posted:

buttbot is yet to be ported to haskell

JawnV6 posted:

yeah string-heavy programthat relies on a perl library SOUND LIKE FUN HASKELLIN

Markov Chain Chomp posted:

bytes are not integers and i will fight anyone who says otherwise

Werthog posted:

this is getting pretty old, this isn't even shaggaring anymore. put some effort into it

tef posted:

Shaggar posted:

lol thats awful

Shaggar posted:

because 1 its ordering on the natural order of butt which isnt the same as a comparator. 2 if you're using a comparator in a bunch of different locations and need to update the logic you do it inside the comparator and you're done. in your shitlanguages you'd have to search and replace for all your compares. 3 the java is far more readable than gross rear end shitlanguages

BonzoESC posted:

does python have both comparator sort and a schwartzian transform sort-by method

BonzoESC posted:

"In Python 2.4 and above, both the sorted() function and the in-place list.sort() method take a key= parameter that allows the user to provide a "key function" (like foo in the examples above). In Python 3 and above, use of the key function is the only way to specify a custom sort order (the previously-supported comparator argument was removed)."

i guess that's "the new way" or maybe "the only way" now

Shaggar posted:

function pointers make me want to vomit

Lysidas posted:

shaggar how do you feel about other callables that aren't technically functions

code:

In [1]: class Adder:
   ...:     def __init__(self, value):
   ...:         self.value = value
   ...:     def __call__(self, other):
   ...:         return self.value + other
   ...:     

In [2]: four = Adder(4)

In [3]: four(5)
Out[3]: 9

ahhh spiders posted:

wahts that have to do with duck typing

ahhh spiders posted:

wouldn't that just be operator overloading

ahhh spiders posted:

the function call operator. calling a function is an operation on the function

Wheany posted:

Oh no, the threads are no longer normalized....

Ridgely_Fan posted:

i remember the pde stuff that was basically breaking stuff down to fourier components after separation of variables, but i would have a hard time solving an ode by hand

irl if i got asked to analyze a system of differential equations i would immediately go to graphical methods

Rufo posted:

calculus is terrible but you need to earn the right to say so

bob arctor posted:

i hated calculus it was the only subject that got in the way of my drinking.

Jonny 290 posted:

[ASK] me about this really depressing theory I have that humans are literally nothing more than bags of water and lookup tables that spend most of their time and resources and life building up one database of i/o but actually only take action when they receive input from another, much smaller and more critical set of inputs