i'm playing with the zx spectrum
i made a circle
[t]http://vgy.me/yAwm3Q.png[/t]
i found a faster but less accurate way to make a circle
[t]http://vgy.me/KRNwdV.png[/t]
i discovered colours
[t]http://vgy.me/P2JZTT.png[/t]
i used colours to make colourful curves
[t]http://vgy.me/5dvV53.png[/t]
i wrote this monstrosity
[t]http://vgy.me/kSbjc7.png[/t]
which produced this
[t]http://vgy.me/jcc996.png[/t]
basic sucks and is slow, so now i'm learning z80 assembly. hoping to make a little game or something with it.
You guys should look at Jane Street hiring process.
Many people say "prepare to die" and "only math gods pass".
Managed to assign appropriate body parts to its appropriate limbs.
Need to make the limb customization more user-friendly later
Which type of lighting/shading is better, if,lets say, the characters are in a digital city in a digital world?
Agree for the boxy one, disagree for the smooth one
[IMG]http://i.imgur.com/b3GIoSY.png[/IMG]
That hiring process sounds like the process for working for the DoD, space industry, or doing defense engineering. I know the army ballistic missile engineering agency does in depth stuff like that, based on the blokes I talked to.
[editline]28th January 2016[/editline]
Hell, sounds like a usual flight physical for the majors
I had a buddy who got accepted for a position with Lockheed Martin but they're so strict about that shit I had to get probed by a FBI agent to make sure he was trustworthy.
Dude FBI agents are serious as shit. I stopped cracking jokes because this agent would just look at me like an idiot.
[QUOTE=Fourier;49626303]You guys should look at Jane Street hiring process.
Many people say "prepare to die" and "only math gods pass".[/QUOTE]
I have a friend that's pretty godlike at math and he didn't pass. They told him that it didn't matter how long he took to come up with an answer, just that he got them right. He got them all right and still didn't pass.
They like to ask mental arithmetic questions like what's the angle between the clock hands at 5:04 (taking into account the hour hand being slightly offset from the 5 position)
[QUOTE=Protocol7;49626881]I had a buddy who got accepted for a position with Lockheed Martin but they're so strict about that shit I had to get probed by a FBI agent to make sure he was trustworthy.
Dude FBI agents are serious as shit. I stopped cracking jokes because this agent would just look at me like an idiot.[/QUOTE]
Yeah, when my friend joined the AF I had to get interviewed since he's in cyber security or something. Another friend got space command or whatever that division is called and I got interviewed again.
I'm terrified of people finding out old shit about me or finding something to DQ me with, since that's the industry I want to go into D:
maybe i should go astronomy instead after all. its pretty much just "hey do you like space science?" "yeah!" "ME TOO! Welcome aboard!"
[editline]28th January 2016[/editline]
not including the 10 years of schooling ofc
[QUOTE=Darwin226;49626941]I have a friend that's pretty godlike at math and he didn't pass. They told him that it didn't matter how long he took to come up with an answer, just that he got them right. He got them all right and still didn't pass.
They like to ask mental arithmetic questions like what's the angle between the clock hands at 5:04 (taking into account the hour hand being slightly offset from the 5 position)[/QUOTE]
what about the fact not all clocks have that difference
otherwise out of curiosity is the answer 2.234?
[QUOTE=Map in a box;49627019]what about the fact not all clocks have that difference
otherwise out of curiosity is the answer 2.234?[/QUOTE]
Something like 128 degrees. Yeah.
[QUOTE=Darwin226;49626941]I have a friend that's pretty godlike at math and he didn't pass. They told him that it didn't matter how long he took to come up with an answer, just that he got them right. He got them all right and still didn't pass.
They like to ask mental arithmetic questions like what's the angle between the clock hands at 5:04 (taking into account the hour hand being slightly offset from the 5 position)[/QUOTE]
That would destroy me, holy fuck.
Ok maybe no, but it would take me 10 minutes to figure it out.
[QUOTE=Fourier;49627436]That would destroy me, holy fuck.
Ok maybe no, but it would take me 10 minutes to figure it out.[/QUOTE]
It's not that bad
Obviously a clock has 360 degrees, an hour hand at 5 would be 150 degrees away from 12 ((5/12) * 360)
A minute hand at 4 minutes would be about 24 degrees ((4/60) * 360)
However the hour hand would be roughly 4/60ths of its way to the 6:00 mark so its angle from 12 is more like 152 degrees (((5/12) + (1/12) * (4/60)) * 360)
152 - 24 = 128
[QUOTE=Protocol7;49627474]It's not that bad
Obviously a clock has 360 degrees, an hour hand at 5 would be 150 degrees away from 12 ((5/12) * 360)
A minute hand at 4 minutes would be about 24 degrees ((4/60) * 360)
However the hour hand would be roughly 4/60ths of its way to the 6:00 mark so its angle from 12 is more like 152 degrees (((5/12) + (1/12) * (4/60)) * 360)
152 - 24 = 128[/QUOTE]
Ah
I was calculating with 2π.
[QUOTE=Protocol7;49626881]I had a buddy who got accepted for a position with Lockheed Martin but they're so strict about that shit I had to get probed by a FBI agent to make sure he was trustworthy.[/QUOTE]
He probably needed a security clearance. Having worked DoD, and having known people who went to other DoD companies (Lockheed Martin included), it's a pretty straightforward process usually. It gets complicated if you need a clearance, especially if you have been married, have foreign contacts, or have any sort of criminal record. My Secret clearance was pretty smooth since I have none of these. I have heard Top Secret is much worse though.
That 4 hour psychological evaluation and personality test is way overkill for any job in engineering.
(5+4/60)/12*360 - (4/60*360)
oh good god sublime text is so much better than matlab's editor and the default theme is actually [I]glorious[/I]. plus its not pure white so it doesn't burn my eyes out at night.
i think i even like this more than n++ tbh
:hammered: :hammered: :hammered:
[t]http://i.imgur.com/bK0ModX.png[/t]
feels bad man
your calendar seems to have a few more days than there are in a month
might be mistaken though
check with a horologist
[CODE]classdef OcT < handle
%{ Octree decompisition in 3D.
% Used to sort n-bodies into subdivisionsfor easier computation, and optimized computation.
% Inspired by Sven's octree system on matlab mathworks
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%{
OT = tree(pts) Generates an octree from an Nx3 array of points,
velocity, or acceleration data
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
The following modifiers can be given to this command with format:
tree(...,'STRING_property',VALUE);
Modifiers are:
bincap = amount of bodies/data allowed in a single bin
maxdepth = how many times a bin can be subdivided into octants
maxL = maximum size of bin walls
minL = minimum size of bin walls
style = 'equal' or 'weighted'. Weighted is more precise but requires
additional computation time. Will produce values that are much more
efficient in terms of space usage. Equal simply divides the parent bin
into 8 equally sized octants, weighted finds most efficient
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Properties:
points = points of properties in decompisition
pointsbin = indices of the bins containing points
bincount = total bins generated
binbounds = BinCount-by-6 [MIN MAX] coordinates of bin edges.
bindepth = The # of subdivisions to reach each bin.
binparent = Indices of the bin that each bin belongs to.
propoct = Name/Val pairs used for creation (see help above)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Properties to add:
%% TO-DO up-chain %%
Sort ALL points based on distance from center of mass M0, order or
index as such, and find relative COM to start problem for cumulative
mass
%%% IMPORTANT %%%
UPDATE
VELOCITY
COM
Force_By
%%% IMPORTANT %%%
Optimize paths between subdivisions?
neighbors within radius
distance k
purge empty octants
%}
properties
points;
pointsbin;
bincount;
binbounds;
bindepth;
binparents = zeros(0,1);
propoct;
end
methods
function this = tree(pts,varargin)
%tree header line
validateattributes((pts),{'numeric','real','finite','nonnan'},...
{'ncols',3},mfilename,'PTS')
numpts = size(pts,1);
this.binbounds=[min(pts,[],1) max(pts,[],1)];
this.points=pts;
this.pointsbin=ones(numpts,1);
this.bindepth=0;
this.binparents(1)=0;
this.bincount=1;
% Allow setting of properties
IP = InputParser
IP.addParamValue('bincap',ceil(numPts)/10);
IP.addParamValue('maxdepth',inf);
IP.addParamValue('maxL',inf);
IP.addParamValue('minL',1000 * eps);
IP.addParamValue('style','equal');
IP.parse(varargin{:});
this.propoct = IP.Results;
% if bin empty/small return
if numpts<2
return;
end
% Start subdividing!
this.preallocatespace;
this.divide(1);
this.deallocatespace;
end
% MATLAB really loves arrays that grow in a loop, so since we hate
% MATLAB lets declare these here and preallocate the space just to piss
% MATLAB off
% BADGER
function preallocatespace(this)
numpts=size(this.points,1);
numbins=numpts;
if isfinite(this.propoct.bincap)
numbins=ceil(2*numpts/this.propoct.bincap);
end
this.bindepth(numbins)=0;
this.binparents(numbins)=0;
this.binbounds(numbins,1)=0;
end
function deallocatespace(this)
this.bindepth(this.bincount+1:end)=[];
this.binparents(this.bincount+1:end)=[];
this.binbounds(this.bincount+1:end)=[];
end
% Function to subdivide bins
function divide(this,startingbins)
for i = 1:length(startingbins)
binno=startingbins(i);
% Prevent dividing beyond the max divide depth from i=1
if this.bindepth(binno)+1 >= this.propoct.maxdepth
continue;
end
% Stop division beyond a minimum
thisbounds=this.binbounds(binno,:);
binedgesize=diff(this.binbounds([1:3;4:6]));
minedgesize=min(binedgesize);
maxedgesize=max(binedgesize);
if minedgesize<this.propoct.minL
continue;
end
oldbincount=this.bincount;
% reasons to divide: bigger than maxsize or more pts than bincap
if nnz(this.pointbins==binno) > this.propoct.bincap
this.dividebin(binno);
this.divide(oldbincount+1:this.bincount);
continue;
end
if maxedgesize>this.properties.maxsize
this.dividebin(binno);
this.divide(oldbincount+1:this.bincount);
continue;
end
end
end
function bindivide(this,binno)
binptmask=this.binpoints==binno;
thisbinspoints=this.points(binptmask,:);
% get old corners and div point
oldmin=this.binbounds(binno,1:3);
oldmax=this.binbounds(binno,4:6);
if strcmp('weighted',this.propoct.style) && any(binptmask)
newdiv = mean(thisbinspoints,:);
else
newdiv = mean([oldmin; oldmax],:);
end
% build new bounds of boxes
minmidmax=[oldmin newdiv oldmax];
newbounds=minmidmax([...
1 2 3 4 5 6;
1 2 6 4 5 9;
1 5 3 4 8 6;
1 5 6 4 8 9;
4 2 3 7 5 6;
4 2 6 7 5 9;
4 5 3 7 8 6;
4 5 6 7 8 9]);
% Determine which points belong in which bins
binmap=cat(3,[0 0 0],[0 0 1],[0 1 0],[0 1 1],[1 0 0],[1 0 1],[1 1 0],[1 1 1]);
gtmask=bsxfun(@gt,thisbinspoints,newdiv);
[~,binassignment]=(max(all(bsxfun(@eq,gtmask,binmap),2),[],3));
% Could use ismember but this tends to be a bit slower
% Make new bins and assign old points back to them
newbininds=this.bincount+1:this.bincount+8;
this.binbounds(newbininds,:)=newbounds;
this.bindepth(newbininds,:)=this.bindepth(binno)+1;
this.binparents(newbininds)=binno;
this.binpoints(binptmask)=newbininds(binassignment);
this.bincount=this.bincount + 8;
end
function binindices = query(this,newpts,querydepth)
% Get bins that contain points given by new points
% USAGE : [vec] = ot.query(newpts)
% Searches octree OT and returns column vector with bin indices containing "newpts"
if nargin<3
querydepth=max(this.bindepth);
end
numpts=size(newpts,1);
newpts=permute(newpoints,[3 2 1]);
binnos=ones(numpts,1)*-1;
binchildren=arrayfun(@(i)find(this.binparents==i),1:this.bincount,'Un',0)';
binisleaf=cellfun(@isempty,binchildren);
ptquery_recurse(1:numpts,this.binparents==0,0)
function ptquery_recurse(newindstocheck_,binstocheck,depth)
% Builds list of points in bins
boundstocheck=this.binbounds(binstocheck,:);
[ptinbounds,subbinnum]=max(all(...
bsxfun(@ge,newpts(:,:,newindstocheck_),boundstocheck(:,1:3)) & ...
bsxfun(@le,newpts(:,:,newindstocheck_),boundstocheck(:,4:6))...
,2),[],2);
if ~all(ptinbounds)
% Case that sometimes happens when at 0 depth
binnos(newindstocheck_(ptinbounds))=-1;
newindstocheck_(~ptinbounds)=[];
subbinnum(~ptinbounds)=[];
end
binnumtoassign=binstocheck(subbinnum);
newindstoassign=newindstocheck_;
binnos(newindstoassign)=binnumtoassign;
% Exit when certain depth reached
if depth>=querydepth
return;
end
% Check which children binned points fall into
[unqbinnum,~,unqgrpnum]=unique(binnumtoassign);
for i=1:length(unqbinnum)
thisptmask=unqgrpnum==i;
if ~binisleaf(unqbinnum(i))
ptquery_recurse(newindstocheck_(thisptmask),binchildren{unqbinnum(i)},depth+1)
end
end
end
end
end
end[/CODE]
yay octree code is vaguely working. as in i can make octree, sort into it, and then index out of it. It still needs a ton of specific functions I have to add though, but I'm still trying to figure out what sort of info I'll need first
Productivity is killed when you import a scale model of the Empire State Building and start challenging yourself to launch yourself off the nearest hill and land on the tip.
Good luck to all the jammers today
Doesn't look very nice, but I'm writing a search engine for a class I'm taking.
Now supports phrase searching !
[IMG]http://i.imgur.com/v7AnODG.png[/IMG]
It's quite slow though. It's searching through an offline version of this wiki: [URL]https://daviswiki.org[/URL]
Edit:
Here's the course book I use. The algorithm I'm using is a tiny bit different.
[url]http://nlp.stanford.edu/IR-book/html/htmledition/positional-indexes-1.html[/url]
[QUOTE=paindoc;49629804]oh good god sublime text is so much better than matlab's editor and the default theme is actually [I]glorious[/I]. plus its not pure white so it doesn't burn my eyes out at night.
i think i even like this more than n++ tbh[/QUOTE]
I keep meaning to pick up Sublime Text since it is really good but I just can't justify the expense.
[QUOTE=Protocol7;49630937]I keep meaning to pick up Sublime Text since it is really good but I just can't justify the expense.[/QUOTE]
I thought Sublime Text was nagware like WinRAR?
[editline]30th January 2016[/editline]
Except that it is socially acceptable to buy Sublme Text.
You should use Atom*
*By should I mean its as good as sublime without the cost, not that sublime is inferior
Something about atom just doesn't sit well with me. And Notepad++ is ass ugly, but at least it's reasonably functional.
[QUOTE=Protocol7;49631149]Something about atom just doesn't sit well with me. And Notepad++ is ass ugly, but at least it's reasonably functional.[/QUOTE]
It's an order of magnitude slower than other editors. It just depends if you have enough hardware so that it's fast enough for you.
I use it as my main editor. You're gonna need this package [url]https://atom.io/packages/keyboard-localization[/url] if you're not from one of the priviledged countries.
Every machine I own has an i7 and SSD so speed isn't really a problem, there are just times when it tries to be smarter than you e.g. with indentations and stuff that I don't really like.
Probably a package out there that would help with that, though, huh.
Are there any reasons to use Sublime instead of Notepad++? I've been using it for ages and I'm not very convinced that Sublime is worth the price.
[QUOTE=Darwin226;49631361]It's an order of magnitude slower than other editors. It just depends if you have enough hardware so that it's fast enough for you.
I use it as my main editor. You're gonna need this package [url]https://atom.io/packages/keyboard-localization[/url] if you're not from one of the priviledged countries.[/QUOTE]
For me sublime takes upwards of 30s to unfreeze when starting up (it seems to just freeze on boot)
Atom takes a second or 2 to startup
[vid]http://files.facepunch.com/ziks/2016/January/29/2016-01-29-1628-04.mp4[/vid]
The game now keeps track of your clipboard history so you can have a palette of parts to build with.
Sorry, you need to Log In to post a reply to this thread.
