The Nineteenth Byte

Helka Homba

2:03 AM

@El'endiaStarman TIL my brain consists of 3.25×10^17 yottabytes of information.

But that's not even golfed...

Geobits

@HelkaHomba I think that's talking about average brains. I'm not sure you fit that.

Ashwin Gupta

github pages isn't updating =/

Helka Homba

@QPaysTaxes Not until the design

Ashwin Gupta

nvm

Ok I need help peooples.

orlp

@HelkaHomba you know the weight and entropy of your head to 2 decimals?

Helka Homba

2:06 AM

@orlp (I just converted 2.6*10^42 bits to yottabytes)

Ashwin Gupta

My Code here. I've got my alien movement equations working/implemented (the green lines are for debugging). How can I set a fix speed for the aliens? I currently am just incrementing the x at a fixed rate but of course this means that the shorter the x, the larger the y, the faster it will go.

Helka Homba

I'm average enough

Ashwin Gupta

wait

orlp

@AshwinGupta make love, not war man...

Ashwin Gupta

dont click

The thing didn';t update

@orlp that isn't an option.

orlp

2:07 AM

@HelkaHomba pls gief new challenge

@HelkaHomba I know you want to post one...

Helka Homba

@Geobits I'll take that as a compliment

Geobits

@HelkaHomba It was deliberately vague, so that's fine. It's what an average person would do, at least.

Helka Homba

@orlp Try to make an omelette from scratch with one hand

El'endia Starman

@QPaysTaxes I'm pretty sure that's actually a main site challenge.

Geobits

Challenge: Given a Chappelle Show screenshot, name the character he's playing. (example given: Ashy Larry)

Ashwin Gupta

2:10 AM

Ok now I got it working

can someone tell me how to fix i?

Geobits

... got it working... problem... does not parse.

@EasterlyIrk OHAI

^^

@Geobits browser?

Chrome

2:11 AM

damn

what does it say?

or do*

Geobits

What are you doing that doesn't work in Chrome?

Helka Homba

@AshwinGupta You just said you got it working

Ashwin Gupta

Its working

for me

Geobits

I'm so lost right now.

Ashwin Gupta

http://niwhsa9.github.io/Space-War/

you went there right ^^???

orlp

2:12 AM

@QPaysTaxes lambda c:[i*a for i,a in enumerate(c)][1:]

El'endia Starman

By the way, I haven't looked at your code, but I'd guess that you event.preventDefault() on every keypress....which means I can't F5 to refresh.

orlp

given an array of coefficients, least significant first

Helka Homba

@AshwinGupta works for me, what's the problem?

Ashwin Gupta

@El'endiaStarman oh yeah, I did do that. You told me too xD

@HelkaHomba okay good its working, i had a problem w/ alien movement

Helka Homba

Oh, they aren't moving

El'endia Starman

2:13 AM

@AshwinGupta Yeah, do it for spacebar and the arrow keys, not for all keys.

Ashwin Gupta

so basically what it is, I calculate a y=mx+b equation based on the line I want the alien to follow

Geobits

I don't know what it's supposed to do/not do, so I don't know what I'm being asked.

Ashwin Gupta

then I increment the x at a fixed rate.

Geobits

Mine follow the green lines and stop at the end.

Ashwin Gupta

But... problem is, that means that when the y is longer, x is shorter, it appears to moves quicker

@Geobits for now its intentionally like that. But read my last message.

How can I get them to move at a fixed rate.

Geobits

2:14 AM

Normalize the movement (x/y vector) and multiply by a set speed.

Ashwin Gupta

Based on the length of the line, not the x distance.

@Geobits so you mean just take the slope and multiply by speed?

Helka Homba

x will be fractionally incremented based on the slope of the line

Geobits

If the slope is normalized, sure. It's been a long time since high school and I don't use slope much any more.

Ashwin Gupta

@HelkaHomba okay I see

@Geobits well its a short time till I go to high school so I don't know much about this stuff :P

Alternative solution I just brainstormed: could I calculate length of the line (using pythagorean) and then do speed based on that?

orlp

@QPaysTaxes 12x^2 -1x is [0, -1, 12]

El'endia Starman

2:17 AM

@AshwinGupta That's what I would've suggested.

orlp

4 mins ago, by orlp

given an array of coefficients, least significant first

Geobits

@AshwinGupta You know how to find distance, right? Divide both x and y by the overall distance, then multiply by your speed.

Ashwin Gupta

@El'endiaStarman I may go w/ that cuz I sorta know what to do.

@Geobits yeah. I didn't know the term but the math makes sense.

orlp

did you listen to what I said?

Ashwin Gupta

Okay i've got some ideas here

I think I know what to do

orlp

2:18 AM

>>> D = lambda c:[i*a for i,a in enumerate(c)][1:]
>>> D([0, -1, 12])
[-1, 24]

Ashwin Gupta

thanks people will report back w/ results

Maltysen

@AshwinGupta what broke?

Ashwin Gupta

@Maltysen besides my head? Nothing really, just implementing movement and running into problems.

orlp

@Maltysen def f(n):k=9**n;return k**-~-~n/~-(k*~-k)%k

Maltysen

@orlp what is this?

orlp

2:19 AM

try that out for some n in your interpreter

El'endia Starman

@QPaysTaxes orlp probably wrote his solution before you updated the challenge. :P

orlp

@El'endiaStarman he updated the challenge?

Ashwin Gupta

@Maltysen I'm using linear (y=mx+b) equations for aliens to follow to a certain random point. At fixed interval, the x is increased and the y is calculated. Problem is, when the y is longer then x, the aliens appear to be moving much faster then the others.

El'endia Starman

> Chat challenge: Given a polynomial function, return its derivative.

orlp

@QPaysTaxes yes it is

Maltysen

2:20 AM

@orlp what does it do?

El'endia Starman

It's pretty standard to have flexible IO on this site.

orlp

@Maltysen did you try it?

Maltysen

yeah

orlp

no pattern detected?

oh I should mention

Python 2, not 3

Maltysen

@orlp o_O

longs don't unpromote?

El'endia Starman

2:22 AM

@orlp Are the precedence rules (and the use of /, not //) the reason it doesn't work in Python 3?

Ashwin Gupta

@Geobits "divide both x and y", so add them and then divide by distance? Or individually divide by distance?

Geobits

individually

orlp

@El'endiaStarman only / vs //

Maltysen

@El'endiaStarman python3 doesn't have a specific long type

@AshwinGupta ah this is trig

instead of using equations of lines, use sin and cos

orlp

@Maltysen but the long part is besides the point :P

Geobits

2:23 AM

@AshwinGupta So if the line was (3,4), you'd have a normalized (0.6,0.8). Then multiply by speed (tweak for best results).

El'endia Starman

@orlp Well, if that's the only difference, then shame on you for using Python 2 instead of 3. :P

orlp

@El'endiaStarman 1 byte man!

Ashwin Gupta

@Maltysen I know some basic trig, I'm like 3/4 through geometry. I can use sin, cos, tan, and the inverse functions

orlp

that's 8 bits

Maltysen

oh fibbonacci

orlp

2:23 AM

do you know how many starving children in africa we can feed using those 8 bits?

Maltysen

@AshwinGupta so instead of using m for the angle, use an actual angle theta, and then do x = cos(theta) and y = sin(theta)

El'endia Starman

@orlp Errr.......hey @Geobits, how many bits do you eat per day?

Ashwin Gupta

@Maltysen Oh got you, I'd have to restructure a bit here. Think the restructure is worth doing? Or should I keep going the way I am?

Geobits

@El'endiaStarman Far more than eight.

Maltysen

@AshwinGupta or if you wanna use m, first do theta = atan(m)

orlp

2:24 AM

@Maltysen you don't think that's curious? that that function calculates fib?

El'endia Starman

@orlp Less than one, apparently.

Ashwin Gupta

@Maltysen yeah that would probably save my lazy butt some work. But maybe I should just restructure

Maltysen

@orlp i hadn't looked at it yet, but now that I have, wtf is going on

@AshwinGupta or make a unit vector

so you don't have to do trig

Ashwin Gupta

@Maltysen don't know what that is =/

Maltysen

take x and y and divide both by sqrt(x*x+y*y)

Ashwin Gupta

2:26 AM

so pythagorean theroem?

Maltysen

but if you're going that far to avoid restructuring, just do the atan thing

@AshwinGupta yeah, that's where it comes from

Ashwin Gupta

Got you

orlp

@Maltysen magic :)

Ashwin Gupta

Hold on, i'm going to look into restructuring here though, it may save me trouble later down the road w/ this.

Maltysen

@AshwinGupta sqrt(x*x+y*y) is the length of the vector, so if you divide by it, it'll make the length 1

Geobits

2:27 AM

@AshwinGupta It's what I've been telling you to do.

El'endia Starman

@orlp That's why I didn't bother trying to understand it. :P

Ashwin Gupta

@Geobits oh now I understand. Sorry the math terms get to my head sometimes :P

Geobits

Lots of games use vectors for movement, so don't feel you need to restructure to use trig.

Maltysen

in fact, trig is for making a vector from an angle and a speed

Ashwin Gupta

Ok ok hold on...

orlp

2:29 AM

@Geobits and all games use linear transformation matrices with homogeneous coordinates, that doesn't mean a beginner should do it that way

Geobits

@orlp Sure, but vectors are drop dead easy :P

Ashwin Gupta

That was alot of stuff. Let me just make sure I got this (and Geo and Malt are on the same page)

Use theta instead of slope.

Then use length of line (pythagorean) to determine speed

Maltysen

@AshwinGupta the only reason for using theta instead of slope is so you can generate them better

Ashwin Gupta

@Maltysen what do you mean generate them better?

Maltysen

cuz theta is in range [-90, 90] m is over R

Ashwin Gupta

2:31 AM

R?

Maltysen

(-inf, inf)

Ashwin Gupta

K I think this may be a good time to remind everyone that i've only taken math through Geometry =/

orlp

he means ℝ

Maltysen

@AshwinGupta here think about it this way, what m will you need to go vertical?

Ashwin Gupta

90

Maltysen

2:32 AM

@AshwinGupta m, not theta

the answer is that you can't / infinity

Ashwin Gupta

Sorry just closed out

lol

undefined isnt it

or 1?

orlp

but he's too cheap to use proper 𝕌ℕ𝕀ℂ𝕆𝔻𝔼

Ashwin Gupta

no horizontal is 1

@Maltysen * ah got it

Maltysen

@AshwinGupta horizontal is 0

orlp

⒧⒤⒮⒫

Ashwin Gupta

2:33 AM

right, jk

Not an idiot, I swear I knew that lol

Maltysen

so, if you're generating theta, you can just make a unifrom random number from -90 to 90

Ashwin Gupta

ahhh

that is probably ALOT better then what I'm doing

Maltysen

m gets more and more steep as it gets higher, and the distribution gets all weird

Ashwin Gupta

I'm randomly picking an x and y.

orlp

beware

if you make a random angle with random length

you don't get an uniform distribution

Ashwin Gupta

2:35 AM

So I can just randomly generate theta, then generate the length I want the line. Then use cos() sin() to get x and y right?

Maltysen

@AshwinGupta just sin and cos

Ashwin Gupta

@orlp ofc not, thats fine IMO

@Maltysen thats what I meant

Maltysen

that's what they do. cos gives you the percentage of an angle in the horizontal direction

and sin gives you the percentage in the vertical direction

then just multiply both by your speed

Ashwin Gupta

ok good :D

orlp

the correct way to generate a random point in a disk

is generate a random angle tau

Ashwin Gupta

2:36 AM

and add?

orlp

generate a random r [0, 1]

Maltysen

@orlp we had a challenge on this or something, didn't we?

orlp

then your x = length * sqrt(r) * cos(tau)

y = length * sqrt(r) * sin(tau)

if you don't do the sqrt you get the problem on the left

Maltysen

@AshwinGupta no you have separate speeds in the x direction and separate speeds in the y direction

orlp

Maltysen

2:37 AM

@orlp that's prolly good for him

Ashwin Gupta

@Maltysen oh right I'm getting rid of the y=mx+b I forgot

Maltysen

i don't think he wants uniform

Ashwin Gupta

Well ty this has been very informational

@Maltysen I don't really care either way honestly.

aditsu

@orlp an easier way is to generate a random point in the enclosing square, and repeat until it's within the disk

orlp

@aditsu fuck ninja'd

what I wanted to say is

Maltysen

2:38 AM

@AshwinGupta haha, k. do the aliens wrap?

orlp

the fastest way to generate random points is generate a random point and reject until the circle

Ashwin Gupta

@Maltysen havent implemnted it yet but maybe

Maltysen

@AshwinGupta then you might wanna only do angles in the range the prevents it from going out of the screen

if you do like 89 deg, they'll leave

Ashwin Gupta

@Maltysen yeah you are right, I'll figure that out

I better start programming this otherwise I'll have to wait until like monday =/

ok brb

orlp

@aditsu def f(n):k=9**n;return k**-~-~n//~-(k*~-k)%k

Maltysen

2:40 AM

right, tommorrow is monday D:

aditsu

insane clown posse - ninja

►

El'endia Starman

I actually came up with both of these solutions to the problem of uniformly sampling a disk to answer an interview question! (It was via email, btw, not in person.) The interviewer said that, apparently, my solutions were creative, and the first time he'd seen a solution to the problem. Apparently, this is a hard thing to do. :P (And I still didn't get the job! >_<)

Alex A.

@Maltysen :(

Ashwin Gupta

@Maltysen yeah except remeber, I have 8 hours of school before I get back to my program...

aditsu

@orlp wat

Alex A.

2:42 AM

@El'endiaStarman Sorry to hear you didn't get it. :( Cool that the interviewer liked your solution though. :)

Maltysen

@AshwinGupta oh, I was merely expressing extreme sadness over that fact

orlp

@aditsu cute, isn't it?

El'endia Starman

@AlexA. Oh well, their loss. :P

Ashwin Gupta

@Maltysen well yes that too

Alex A.

@El'endiaStarman Definitely

Ashwin Gupta

2:43 AM

@El'endiaStarman did you find something else yet* or still looking?

aditsu

@El'endiaStarman "the first time he'd seen a solution" -> you clearly didn't fit in with the rest of the team

El'endia Starman

@AshwinGupta It's been something like 9 months since then and I'm still looking. :(

Ashwin Gupta

@El'endiaStarman aw sorry dude, don't worry, best always comes last.

El'endia Starman

@aditsu ...I like that way of looking at it. :P

Alex A.

I currently have a job but I've been looking for a different one for almost year and a half. ._.

Ashwin Gupta

2:44 AM

@El'endiaStarman yeah close-minded people have difficult accepting genius sometimes...

El'endia Starman

@AshwinGupta The Marriage problem would disagree with that. :P

Ashwin Gupta

@El'endiaStarman :D LOL

El'endia Starman

@AshwinGupta I'll bet the close-minded person was not the interviewer, but his boss (or their boss, etc.). :P

Ashwin Gupta

Stupid question: How do I generate a random number between -90 and 90 in js?

Math.random doesn't let me do that?

@El'endiaStarman normally the case honestly.

El'endia Starman

@AshwinGupta You can do Math.random()*180-90.

Ashwin Gupta

2:47 AM

@El'endiaStarman probably for the best you didn't get stuck there though. It would suck to be stuck in a job where you aren't allowed to use any creativity.

Maltysen

@AshwinGupta take math.random, * 180 - 90

Ashwin Gupta

@El'endiaStarman right duh. Ty

Maltysen

ninja'd

aditsu

@orlp so that's a Fibonacci function? why did you want to show it to me?

El'endia Starman

@AshwinGupta Yeah, perhaps.

orlp

2:48 AM

@aditsu yes, it's fibonacci, and I just found it intriguing

El'endia Starman

I was about to ask if it was alright to convert floats to ints when its value is integer, given that there are no operations that act differently for ints and floats, when I remembered that all three spreadsheet programs I've used (Microsoft Office, Open Office, Google Docs) do that, and it's annoying for aesthetic reasons.

Maltysen

3:02 AM

we don't have mathjax, right?

@El'endiaStarman using the format menu doesn't work?

orlp

$f(x) = x^2$

@Maltysen math.ucla.edu/~robjohn/math/mathjax.html

put "start chatjax" on your bookmarks

Maltysen

@orlp I was talking about in an actual challenge

orlp

images

unicode

super/subscripts

Maltysen

I'm sure code blocks are fine

Downgoat

@Maltysen Who needs MathJax, when you have monospace, ASCII, and enough determination

El'endia Starman

3:12 AM

@Maltysen Yeah, there are ways to format such numbers properly, but I was thinking of the default behavior.

@QPaysTaxes Agreed.

Incidentally, I also have ChatJax/MathJax. :D

Downgoat

strawpoll.me/10042568

Maltysen

@Downgoat I don't have enough determination

@Downgoat the question isn't stacksnippets vs. mathjax though. its mathjax vs. using $() instead of jQuery() in stacksnippets

Downgoat

Solution is to not use jQuery

or to fix MathJax

Why doesn't SE fix MathJax??

brb, writing rant on meta

Maltysen

@Downgoat you can still use jQuery, just not its alias

Downgoat

jQuery is bad you should never use it

jQuery is horrible

jQuery is worse than JavaScript

Alex A.

3:15 AM

MathJax also broke search

Downgoat

fuck, precedences on Cheddar broke ;______________;

this is veeerrrryyyy bad

where?

Maltysen

@QPaysTaxes are you trying to make an rng out of humans?

Downgoat

@QPaysTaxes ಠ_ಠ

what is "Blame"?

Does it let me find who to blame for cheddar breaking?

Maltysen

@Downgoat yeah, all the entries in there will be you, and it'll be correct :P

Downgoat

ಠ_ಠ

@MarsUltor ShuntingYard in Cheddar broke

Also, I made Cheddar's readme more fancy

Maltysen

3:22 AM

why does it say build:passing?

Downgoat

@Maltysen because the build is passing

Hello @Doorknob!

Maltysen

@Downgoat I thought you just said something was broked ;P

Downgoat

@Maltysen yeah.

Maltysen

@Downgoat then why is it passing?

Downgoat

because the code isn't borked, just the functionality

Maltysen

3:24 AM

oh

Downgoat

I don't have tests set up though, I really need to do that

Right now travis is just building and pushing to heroku

Doorknob

@EasterlyIrk Please stop pinging me for no useful purpose.

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Why did I come to Sandbox? I have a very specific challenge, and I wanted to see if it was too specific. The challenge is to output "Valdosta ACM" using the shortest number of characters with the BrainF**k programming language. I've noticed it isn't the norm to specify a programming language...

Downgoat

3:39 AM

O_o

@QPaysTaxes Doesn't google customize search results for you...?

New Main Posts

1

Q: Taylor Series of a Function with Periodic Derivatives

Taylor series are a very useful tool in calculating values of analytic functions that cannot be expressed in terms of elementary functions, using only information about that function at a single point. In this challenge, you won't be actually doing any math with them, but merely making string re...