Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Arx]

You won't believe guys but some say programming languages are simply a form of representation of algorithms, and in order to become a programmer I should focus on learning algorithms rather than on learning a programming language. Are they completely right? Whether they are or not, where do I start from? What books (or another resources) on algorithms would you guys particularly suggest and what algorithms is it necessary to know for every programmer?

They're right that programming languages are just a way to represent algorithms, but I suggest learning a language anyway. It's one of the easiest and most hands-on ways to get practice with algorithms and algorithm development.

By and large, you should work on being able to find an algorithm for a given problem rather than memorising them. Most must-have algorithms will be a standard library in most languages, so I wouldn't worry overmuch about that.

[RoguelikeRazuka]

Spoiler (click to show/hide)

#include <iostream>
#include <ctime>
#include <cstdlib>

using namespace std;

int main(){

   int userChoice, compChoice;

   int rock = userChoice = 1;


   srand(time(0)); //RNG here.

   cout << "Welcome to the game of games.\t" << endl
      << endl << endl;


   cout << "1. is for rock, 2 is for paper, and 3 is for scissors. Now pick a number.\t" << endl
      << endl << endl;

   cin >> userChoice;

   cout << endl << endl << endl << "You have picked " << userChoice << " .\t" << endl
      << endl << endl;

   if (userChoice == 1){
      cout << "Rock." << endl << endl
         << endl;
   }
   else if (userChoice == 2){

      cout << "Paper" << endl << endl

         << endl;
   }
   else if (userChoice == 3){

      cout << "Scissors" << endl << endl

         << endl;
   }

   
   compChoice = ((rand() % 3) + 1);




   while (userChoice == compChoice){
      cout << "You have tied with your opponent. Try again. \t" << endl << endl << endl;

      cin >> userChoice;
      compChoice = ((rand() % 0) + 1);
      //When opponent pick the same as player.
   }

   while (userChoice >= 4 || userChoice < 0){
      cout << "Have you no shame! You didn't pick 1-3. Try again." << endl
         << endl << endl;

      cin >> userChoice;

      //In case people decide to use negatives or numbers higher  than 3.

   }


   //Below are the winning conditions.

   if (userChoice == 1 && compChoice == 2){
      cout << "Your opponent picked: " << compChoice << endl
         << endl << endl;
      cout << "You lose" << endl << endl << endl;
      //user rock comp paper
   }

   else if (userChoice == 1 && compChoice == 3){
      cout << "Your opponent picked: " << compChoice << endl
         << endl << endl;

      cout << "You win!" << endl << endl << endl;
      // user rock comp scissors 
   }
   else if (userChoice == 2 && compChoice == 1){
      cout << "Your opponent picked: " << compChoice << endl
         << endl << endl;
      cout << "You lose!" << endl << endl << endl;
      //user paper comp rock
   }
   else if (userChoice == 2 && compChoice == 3){
      cout << "Your opponent picked: " << compChoice << endl
         << endl << endl;
      cout << "You lose!" << endl << endl << endl;
      //user paper comp scissor
   }
   else if (userChoice == 3 && compChoice == 1){
      cout << "Your opponent picked: " << compChoice << endl
         << endl << endl;
      cout << "You lose!" << endl << endl << endl;
      //user scissors comp rock
   }
   else if (userChoice == 3 && compChoice == 2){
      cout << "Your opponent picked: " << compChoice << endl
         << endl << endl;
      cout << "You win!" << endl << endl << endl;
      //user scissors comp paper
   }
   


   system("pause");

   return 0;

}

Found a problem that after a tie it won't check for incorect responces (Because it's already been done). Something else happens when I swtich them around.

Another one is that if I switch the two loops around then something happens that after I've tied and then input wrong statements that my program will accept a wrong statement before it simply just dosen't declare a winner. This is C++ btw.

Try this:

   cin >> userChoice;
   compChoice = ((rand() % 3) + 1);

   while (true)
   {
      if (userChoice < 1 || userChoice > 3)
      {
         cout << "Have you no shame! You didn't pick 1-3. Try again." << endl << endl << endl;
         cin >> userChoice;
      }
      else if (userChoice == compChoice)
      {
         cout << "You have tied with your opponent. Try again. \t" << endl << endl << endl;
         cin >> userChoice;
         compChoice = ((rand() % 3) + 1);
      }
      else
      {
         break;
      }
   }

Hope it helps!

[3man75]

what does the else part work? I've never seen break as a piece of code.

[RoguelikeRazuka]

what does the else part work? I've never seen break as a piece of code.

It just escapes the while statement if the correct (i.e not below 1 or above 3 and not equal to the computer's choice) number has been got and proceeds towards the next statement.

[TheBiggerFish]

break;
will remove you from a looping construct regardless of the continuation condition.

[Gatleos]

I just solved a programming problem with a combination of template metaprogramming, polymorphism, and bound function objects. I suddenly feel so vindicated.

[TheBiggerFish]

I just solved a programming problem with a combination of template metaprogramming, polymorphism, and bound function objects. I suddenly feel so vindicated.

I have no idea what any of this means.  Nonetheless, congratulations!

[TheDarkStar]

I just solved a programming problem with a combination of template metaprogramming, polymorphism, and bound function objects. I suddenly feel so vindicated.

Out of curiosity, what was the problem?

[Gatleos]

I just solved a programming problem with a combination of template metaprogramming, polymorphism, and bound function objects. I suddenly feel so vindicated.

Out of curiosity, what was the problem?

I was trying to find a way to elegantly represent animations for graphical primitives, an animation being broadly defined as "a property changing over time". So a single object representing something like a square being drawn on the screen could have individual animations affecting its position, rotation, scale, color, etc. without having to create a huge mountain of classes to do it.

My solution was to define three classes, derived from a base Animator class. The first is called anim::Scalar, and it is constructed using a function object that takes a single float value as a parameter. There are two more, anim::Vector and anim::Color, which take coordinate and color arguments, respectively. Then I pump the Animators full of KeyFrame objects (templated with the type of value they animate), and cycle through those while interpolating the values (float, float vector, and color).

That probably only made like 40% sense, but you get the basic idea. Honestly I'm surprised it worked out as well as it did.



Which reminds me, I have a question: does anyone here have opinions on data serialization formats? I'm planning on storing and loading those animation objects I described as files, and I've been trying to find a compact, non-flat data representation format to use. Is there something compact and not necessarily human-readable that has a hierarchical format like XML or JSON?

Though I get the feeling human-readability is still a good idea even if it's just for program input and output.

[TheBiggerFish]

So...Graphics, the bane of my existence?  (I seriously seriously hate rendering things.  Java.)  Yup.  You, sir, have done Great Things today.

[Skyrunner]

I very much like JSON because even if it is human-readable, printing it with no spaces might as well not be human-readable It is fairly compact, too, and you could compress it for further space-saving. And the rules are super simple and terse. XML has a toooon of tag (and flexibility) but in almost all cases you don't actually need XML's RADICAL INFINITE CHOICE, and JSON suffices, with its brief syntax of []{},":.

[jaked122]

Really? I love the Java graphics primitives, very simple to use for a lot of things. Just doesn't scale to animations very well, no tweening, difficult clipping, etc. Also Java is just too pedantic for that IMHO.

That solution for representing animations though... I'm not sure about the need for bound function objects, that seems a bit less efficient to me, but I've never read the documents in the standard(I'd assume c++), but despite the efficiency of your templates, bound functions aren't necessarily optimized very well.

I might very well be wrong, especially if the template parameters include the bound function, that would be very nice.

[Gatleos]

I might very well be wrong, especially if the template parameters include the bound function, that would be very nice.

No, you're right. It does use more cycles, since it needs to lookup a function pointer whenever it updates. It just makes such a tiny difference in the processor load that the convenience was worth it. It was either that or define individual classes for animating position, scale, rotation, origin, fill color, outline color, not to mention different types of objects, some of which don't have setter functions. The way I did it, just a few Animator classes handle everything.

I was worried about the loss in speed at first, then I noticed I could update about 30000 animations per frame while still running at 60fps.

[monkey]

http://msgpack.org/ is a nice lib for binary JSON, someone made a nifty tool to convert from text to binary and back, look in the Shell implementation.
And animation data is kind of hard to edit by hand, so you are not losing much by using binary files.

[jaked122]

Why would you use binary json? The whole point is that it can be loaded easily by either javascript or python.

Besides, if performance matters, but you can't have a large json file, then I'd use zlib to compress it with a deflate stream, which works fairly well and won't slow your modern machine down too much.