|
Thread Rules
1. This is not a "do my homework for me" thread. If you have specific questions, ask, but don't post an assignment or homework problem and expect an exact solution.
2. No recruiting for your cockamamie projects (you won't replace facebook with 3 dudes you found on the internet and $20)
3. If you can't articulate why a language is bad, don't start slinging shit about it. Just remember that nothing is worse than making CSS IE6 compatible.
4. Use [code] tags to format code blocks. |
On November 19 2010 00:35 KaiserJohan wrote:Hmm, anyone know if there really is no way of getting the size of a class instance in C#? sizeof only returns the size of the pointers, and Marshal.Sizeof does not work on class instances.  It kinda makes the #1 solution to object-to-byte streaming hard :p Anyone know any other way of getting size of a class instance in C#? (imho it feels like a major limitation if its not possible)
Why does it make that hard? Write out the length of the serialized object to the stream before you write the object. Read the length, and then read the object. Problem solved.
No, there isn't a framework method for crawling all the references and getting the size like that. But that wouldn't be what you want, anyway. You want the length of the serialized representation, and if you're using any reference types that you didn't write yourself, then the serialized length might well be different from the size in memory.
|
Anyone good with NP-C reductions want to lend me a hand here?
Tree Reconstruction Problem:
Given a collection of trees: T_1, T_2, ..., T_n
and a positive integer m
Decide whether there exists a tree T of size at most m, such that T_1,...T_n are all isomorphic to a subtree* of T.
Prove that TreeRec is NP-complete
+ Show Spoiler [*Subtree isomorphism] +
Given H and G, H is isomorphic to a subtree in G iff
For function h: V(H) -> V(G) (vertices of a graph)
For every edge (u, v) in H, there exist vertices h(u) and h(v) joined by an edge in G.
Anyway, I tried these reductions:
Subgraph isomorphism (I couldn't really link trees with graphs, not to mention N trees)
Clique (I tried linking "m" to a clique of size "m", didn't work out; but I think this is the closest one I have)
3-SAT (I got stuck)
m-SAT (I got even more stuck)
I e-mailed my TA for some help, and they suggested that I try 3-colour, and I got no luck with that either.
Given the decision problem, it takes "m" as an input, the reduction problem should also use "m" somewhere (that's why I tried m-SAT), and I don't really understand where I could go with 3-colour.
Help/hints would be greatly appreciated...
|
On November 14 2010 01:24 Osmoses wrote:Hey gang, I've been working on a game recently, a grid based strategy game focusing on good performance, but now I've started working on pathing, since I want something like workers responding to clicks, and the performance is taking a hit. If it's just corridors, it's easy, big O of 1 most of the time, but once I start working with open spaces, it gets alot more complicated trying to find the shortest path with the least amount of work. I'm trying out a couple of different methods of going about this right now, but in the meantime I thought I'd ask if anyone can recommend a good book or something concerning 2d pathing? Below is a screenshot of a potential stage, with a example start and end marker placed. ![[image loading]](http://imgur.com/7TPoT.png)
If your A* is too slow then heuristic may need tweaking, you may need some common sense performance tweaks, and above all you want to break down the problem into easy steps. Try double-ended searching to waypoints along the path and store these results somewhere. Also, don't waste your time searching on trivially small spaces - you can build a hash table of precomputed paths based on the pattern of impassable terrain (like, 3x3 or 4v4).
Even more badass is to just guess the answer.
|
Germany2896 Posts
@KaiserJohan
If you want to serialize .net objects why don't you an existing serializer. And writing a serializer is easy compared to writing a deserializer.
* WCF is can serialize objects in order to send them over the network
* There is BinarySerializer which can serialize most object graphs
* There are several Proto-Buf serializers which are really fast and the serialized data they produce can be (if you create a protobuf description file) used from other languages or different versions of your program.
* There are XmlSerializer and some Json serializers, but they are probably not what you want.
And register or stackoverflow. Very good site for concrete programming questions with definite answers, but bad for discussions.
|
France1920 Posts
|
I have a basic C# question thats really giving me a lot of trouble. Note, this the first assignment or program I've ever attempted to use C# for. The point of the assignment is to teach the chain of responsibility design pattern.
Basically, the code looks like this.
Abstract_Handler.cs
+ Show Spoiler +
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
namespace CIS476_Project3_Example1
{
public class Abstract_Handler
{
protected Abstract_Handler next;
public Abstract_Handler()
{
next = null;
}
public void addLink(ref Abstract_Handler n)
{
if (next != null)
{
next.addLink(ref n);
}
else
{
next = n;
}
}
public virtual void handleRequest(ref Request x)
{
next.handleRequest(ref x);
}
}
}
USD_Handler.cs
+ Show Spoiler +
namespace CIS476_Project3_Example1
{
public class USD_Handler : Abstract_Handler
{
public const float EURO_TO_USD_RATE = 1.04f;
public override void handleRequest(ref Request x)
{
x.value_1 *= EURO_TO_USD_RATE;
if ( x.CurrencyType != 0 )
{
handleRequest(ref x);
}
}
}
}
Inside of Main( ) I call the following code:
+ Show Spoiler +
/* Instantiate Handlers */
USD_Handler USD = new USD_Handler();/* Root */
CAD_Handler CAD = new CAD_Handler();/* Link 2 */
AUD_Handler AUD = new AUD_Handler();/* Link 3 */
/* Create Chain of Responsibility */
USD.addLink( ref CAD );
USD.addLink( ref AUD );
Now when I build, I get an error flagging those last 2 signs that says soemthing along the lines of Abstract_Handler.addLink(ref Abstract_Handler) has the wrong parameter, even though USD, CAD, AUD Handlers are all subclasses of the abstract handler. Am I doing something wrong with inheritance or virtual function overloading? Am I using references correctly? At this point I really have no idea. I tried C# because of how easy GUI work seems to be.
For reference, here's my entire program written in C++.
http://codepad.org/KJ87r5fs
|
I'm fiddling around with it (my C# is so-so). It'd help if you posted your Request class (and the other classes so I can test w/ them).
It looks like you're using ref where you don't need to be or at least wrong somehow. Not sure atm. If I could just figure out why I'm throwing a NullReferenceException after finishing the last conversion I'd be set...
http://codepad.org/7rZRxHl9 Seems to be working.
As I understand it, objects passed around in C# are inherently passed by reference. C# also doesn't have pointers.
|
On November 24 2010 02:43 Craton wrote:
As I understand it, objects passed around in C# are inherently passed by reference.
Thanks a ton for your help man. Cleared up quite a few things. After working almost entirely on embedded C code the past two years, I'm starting to overthink some things in these higher level languages it seems :p
|
On November 24 2010 02:43 Craton wrote:
As I understand it, objects passed around in C# are inherently passed by reference. C# also doesn't have pointers.
Nitpicker's corner: C# does have pointers, but they're not useful for this. C# objects aren't exactly passed by reference in the same way that C++ functions can pass by reference:
void swap(int &x, int &y) { // in C++
int temp = x;
x = y;
y = temp;
}
void swap(MyType x, MyType y) { // in C#
MyType temp = x;
x = y; // NOT doing the same thing!
y = temp; // caller's reference will not be swapped!
}
It would be accurate to say that C# passes object references by value. When a function receives a reference type as a parameter, it receives a new reference to the same object. If the object is mutable and you change fields on that object, they'll be visible to anyone else who has a reference to the same object. (This is the case with your Request object.) However, you don't have the same reference that the caller has, and you can't make the caller's reference point to a different object, or to null.
So that's what the ref keyword is for. If you use ref parameters, you're actually passing by reference in the C++ sense.
void swap(ref MyType x, ref MyType y) { // this works like C++
// ...
}
|
It would be accurate to say that C# passes object references by value. When a function receives a reference type as a parameter, it receives a new reference to the same object.
What is the difference between the two?
|
public static void Main(string[] args)
{
object alpha = new object(),
beta = new object();
// at this point in time, we have two references, alpha and beta.
// each one is pointing to some object somewhere on the heap
SetToNullA(alpha);
SetToNullB(ref beta);
// alpha is not null. beta is null.
}
static void SetToNullA(object x)
{
// we passed the reference to alpha by value, making a copy of it.
// now we have THREE references, alpha, beta, and x.
// x is pointing to the same object as alpha -- it's a copy
// of alpha. we could change things about that object using x.
// but x is not alpha. if we set x to null,
// that does not mean alpha is null. this function is a no-op
x = null;
}
static void SetToNullB(ref object y)
{
// we passed the reference to beta by reference.
// y is the same reference as beta. if we set y
// to null, then beta is null, too.
y = null;
}
|
On November 24 2010 09:44 Craton wrote:Show nested quote +It would be accurate to say that C# passes object references by value. When a function receives a reference type as a parameter, it receives a new reference to the same object. What is the difference between the two?
When you send an object in C# by default it is like a pointer in C++
void mowlawn(LawnMower* mower) c++
void mowlawn(LawnMower mower) c#
Those are equivalent. In both cases setting mower to something else in the function has no effect on the original pointer to mower that was sent in.
void mowlawn(LawnMower& mower) c++
void mowlawn(ref LawnMower mower) c#
And these two are the same thing. If you change what mower points to in the function, the caller mower pointer is changed.
|
On November 22 2010 19:03 HaN- wrote:![[image loading]](http://imgur.com/N4vDL.jpg)
Hahaha, you sir win the intarwebz
|
Germany2896 Posts
On November 24 2010 13:21 HowitZer wrote:
void mowlawn(LawnMower& mower) c++
void mowlawn(ref LawnMower mower) c#
And these two are the same thing.
Those two are completely different. The C# code is more like
void mowlawn(LawnMower** mower)
"What is the difference between the two?"
pass by reference would imply that the variables you normally work with have value sematics. But they are references themselves. And when you pass the reference to a function it gets copied, just like a value would be copied.
|
Just wanted to let you guys know that these days I finally learned about www.stackoverflow.com and my life changed since then. That's the best website in the history of the internet. If you code on any level and haven't been there yet, you should.
The community is awesomely well organized, it's impressive. TL could get some ideas from there. Their system really works well.
|
Hi guys first time poster here, I have a question about creating a square wave table in JAVA
Ive found out that the formula for it is amplitude at max (127) for half the cycle and then zero for the other half \
so the wavetable is only 2 elements
my given frequency is 440 hz and my sample rate is 44100
can anyone help me out for creating a private instance variable for an array of integers that represent a wave table? the consturcter has to be the code for creating and populating the array
|
EDIT: Nevermind, my bad >_>
|
I really screwed this semester up, and if worst comes to worst I may even have to retake one of my comp sci classes. I'll probably be dropping from my 3.25 gpa to a 2.6 -- honestly haven't felt this terrible for a long time.
Am I screwed now for finding an internship? Does gpa play a big role in finding work? I'm headed into my 2nd semester of junior year and kind of feel like i'm way behind the pack. I'm straight As from here on out, swear to god -- but has the damage already been done?
|
Competence plays a big role in finding work; there is a huge, perpetual shortage of competent programmers. If you can point at reasonable-quality open-source work that you have done you are automatically sitting on top of the pile, unless you're trying to intern at Google. Worry about your Github exclusively instead of worrying about your GPA.
|
hey guys. i'm having a problem right now about the correct algorithm to use in computing exponent. we all know that the basic way to do it is by iteration ( 2^4 = 2 * 2 * 2* 2), but i want to do it recursively so that i will execute faster.
for example, if my exponent is an even number, i have to divide it by 2 until it will not be divided anymore, it will look like this
+ Show Spoiler + 2^e = 2^e/2 * 2^e/2
or
2^8 = 2^4 * 2^4
2^4 = 2^2 * 2^2
2^2 = 2^1 * 2^1
or
2^10 = 2^5 * 2^5
2^5 = 2^4 * 2
2^4 = 2^2 * 2^2
and if my exponent is an odd number, it will look like this
+ Show Spoiler + 2^e = 2^(e-1) * 2
or
2^5 = 2^4 * 2
2^4 = 2^2 * 2^2
2^2 = 2^1 * 2^1
i did my best, and my head is really aching right now. i hope someone will be able to help me.  here is the code that i made.
else if (e%2 == 0){ // if the exponent is even
for (i = 0; i < e; i++){
e = e/2;
if (e%2 == 0){
for (i = 0; i < e; i++){
res = res * b;
res = 2*res;
}
}
else if (e%2 == 1){
z = e-1;
for (i = 0; i < z; i++){
res = res * b;
res = 2*res;
}
res = res * b;
}
}
printf("%d\n\n", res);
}
else if (e%2 == 1){ // if the exponent is odd
for (i = 0; i <= e; i++){
e = (e-1) * e;
e = e/2;
res = res * b;
}
printf("%d\n\n", res);
}
thanks guys!
|
|
|
|
|
|
EPT
