Can anyone tell me how does range based for loop works with array of pointers Suppose i have an array of pointers which has pointers pointing to integers. for(auto a:array) *a = 0; All the elements will become zero. But by default range based for loop(without reference) passes copies right. Then how did it affect the real ones?

Can anyone tell me how does range based for loop works with array of pointers Suppose i have an array of pointers which has pointers pointing to integers. for(auto a:array) *a = 0; All the elements will become zero. But by default range based for loop(without reference) passes copies right. Then how did it affect the real ones?

And by the way, "copy by default" is result of using auto here, because auto doesn't deduce references. Range based loops provide references, but auto drops it. You can use auto& here, or specify the type.

how do i detect if a value is less than X, for example, given an array of {2, 4, 6, 8} how do i know if my value went from: 10 to 8 (max) or 7, 6 (min) to 5, 2 (min) to 1, assuming ranges of min 2 max 4, min 4 max 6, min 6 max 8

cant

the array is to be interpreted as a sequence of min/max

i can detect if the value goes up (eg 2 to 3, 4 to 7) by just doing if (val > min_ && val < max_ && cached_val_max != max_) but i doing the same for down doesnt work

I started learning c++ through youtube and some materials but i am not improving, is there any other way.

I started learning c++ through youtube and some materials but i am not improving, is there any other way.

Hi, I have GUI, which is sending from multiple threads some messages to message queue. Then engine reads that messages and sends some replies, or data, errors, statuses in other message queue. Messages, sent from GUI have unique ID, so when Engine send reply back, it uses this ID. How can i handle Engine replies? So, i need to identify to which GUI request, Engine sends reply. Engine can send not only replies, but also as i mentioned, some statuses, errors... I need to be able to handle this in GUI too. It mostly about DESIGN. What would you recommend to achieve what i need?

Are the GUI threads doing similar work i.e are they just sharing the workload to make it faster or are they meant to do different tasks that are independent of each other?

They are intended to to different tasks. It's like a GUI threads. One thread can operate over user input, while other is drawing something (just example) But for the future would be great to include solutions for both this cases,

In that case have a queue for the engine where these GUI threads will be posting their messages. The engine can process them sequentially or even have multiple threads of its own reading off this queue. Each GUI thread (since it is independent) can have a separate queue of its own where the Engine can post it's response. Have a register mechanism on the Engine where these GUI threads can register their queues. The Engine can return a unique ID to these GUI threads which can be used to identify the queues. When a GUI thread posts a message into the engine, it should include this unique Id in the message. This is the simplest way to achieve what you intend to do.

Ok then have another thread in the GUI library that reads the engine responses from the queue meant for GUI and have it pass the message to the correct GUI thread using a callback mechanism. In this case the GUI threads don't have to register with the engine.

Yeah, i have already sender and receiver threads in GUI, Can you, please, describe with more details your vision about callback mechanism in current case?

Callback mechanism is a popular concept in multithreading and elsewhere too. There are many resources for it online. Would be more easy to learn about it that way rather than posting all the details here

I understand what you are talking about. I'm wondering about proper way to integrate them for my issue. Example: GUI sends from thread1, request1, then start waiting for reply1. In the same time GUI sends from thread2, request2, but it don't wait to reply (but as soon as it arrives from Engine, GUI should for example notify user) Then Engine, pops, let's say, request1, but before sending reply1, it sends others replies(like result1, result2, result3...) after which it finally sends reply1. Same can happens with thread2. So, i'm trying to say, that i don't need single request-response. I need also to be able to operate over others replies from Engine.

Forget the engine. Assume the engine sends responses in random order and that it doesn't necessarily send responses in the same order as requests which is likely what will happen in real world. The GUI Task threads can either fire a request to the engine and wait for it's response or they may send an async request and process the response when it arrives. In the latter case you need a store and forward mechanism (the GUI thread handling GUI queue should store the responses from the engine and forward them to the GUI thread meant to handle it) for the responses to be processed. A better way instead would be if you could use a promise/future combination with the future holding all the information that the GUI thread needs (i.e. the response details from the engine) and the GUI thread will maintain the future objects for all the requests that it sent out. The GUI thread responsible for the GUI queue in which the engine responses will be stored will pick these responses (in the case of promise/future combo this will just a request response id which is just a notification from the engine that the request has been processed. The response details will be in the future object associated with the promise) and then use the callback of that GUI thread. This callback will typically set an interruptible state informing the GUI thread that there are responses from engine that need to be processed. The GUI thread after finishing what it has done (or if the interrupt is of higher priority stop doing what it is currently doing) andcould check the response id and use the corresponding future object to retrieve the responses for the engine, process it and then go back to doing what it was doing earlier. If there are many responses, then it could process all of them based on their priorities.

#include <bits/stdc++.h> using namespace std; void func(string ans,int n) { if(ans.size()==n) { cout<<ans<<" = "<<endl; return ; } ans+="0"; func(ans,n); ans.pop_back(); ans=+"1"; func(ans,n); ans.pop_back(); } int main() { int n; cin>>n; string ans=""; func("",n); }

If someone know anything about Qt please send me a message. I have many questions!

https://youtube.com/user/VoidRealms This guy knows everything about Qt

May I please know the reason if I enter the input as alphabets then why it will give the result as positive ? (Everything else works fine.) Even though I put negative sign (-) prefix to alphabet then it will execute result like "Given number is positive".

Yeah..if I enter the positive or negative number then I am getting the correct corresponding results.

Ah, I've misread it, what do you mean by "alphabet"? You mean non-number, like "a"? In this case it's an error case and scanf should return some special error value (and you should not rely on the value of num). I don't remember what value is returned from scanf in this case..

Ah, I've misread it, what do you mean by "alphabet"? You mean non-number, like "a"? In this case it's an error case and scanf should return some special error value (and you should not rely on the value of num). I don't remember what value is returned from scanf in this case..

Yeah...Got it, but why it taking alphabet as positive ? Why don't it print as negative ?

minus doesn't make sense with not numbers, I guess it just gives some garbage value. In my case it gives zero always, doesn't matter what symbols I enter.

#include <stdio.h> /* finding prime factors of a number n*/ int main(void) { int n, i, j, count = 0; printf("enter n\n"); scanf("%d", &n); printf("prime factors of %d are\n", n); for (i = 1; i <= n; i++) { for (j = 1; j <= i; j++) { /* prime has only two factors, count = 2 */ if (i % j == 0) count++; } /* if prime and also a factor of n */ if (count == 2 && n % i == 0) { printf("%d\n", i); } } return (0); } Pls can someone help with this code? It is not giving right output although it compiles

This is inefficient way of checking prime factors. Anyway your problem would be resolved by adding the line count =0 as the first statement of your outer for loop i.e. before the inner for loop

Prime Factorization is NP Hard so there is no known polynomial time algorithm. The easiest one in your case would be Pollard's rho algorithm (you can search for it online). If you want to make your algorithm more efficient, you can skip even numbers (by incrementing your inner loop counter by 2) and also by iterating your inner loop till j*j < i insead of all the way till i

If I skip even numbers, 2 will be affected right? But 2 is a prime number

You can skip all even numbers except 2...check for divisibility by 2 outside the inner for loop and start iterating from 3 and skip even numbers

Prime Factorization is NP Hard so there is no known polynomial time algorithm. The easiest one in your case would be Pollard's rho algorithm (you can search for it online). If you want to make your algorithm more efficient, you can skip even numbers (by incrementing your inner loop counter by 2) and also by iterating your inner loop till j*j < i insead of all the way till i

You can make it even more efficient by removing the isPrime check and while it's in the NP-class, integer factorization is not known or believed to be NP hard.

He wants prime factors. So how can he skip the isPrime check? And integer factorization is believed to be NP Complete even though there is no known proof. I should have said NP Complete

by something like this, once you know the number is not divisible by any number smaller than c but is divisible by c, c must be a prime, e.g. void printPrimeFactors(int n) { for (int c = 2; n != 1; c += 1) { while ((n % c) == 0) { printf("%d ", c); n /= c; } } }

by something like this, once you know the number is not divisible by any number smaller than c but is divisible by c, c must be a prime, e.g. void printPrimeFactors(int n) { for (int c = 2; n != 1; c += 1) { while ((n % c) == 0) { printf("%d ", c); n /= c; } } }

This is too expensive and I don't understand

For ex if you are finding the prime factors of 60 You will do this: 60%2 = 0 So 2 is a prime factor. Then 60/2 = 30 30%2 = 0 So 2 is another prime factor. Then 30/2 = 15 15%2 != 0. So 2 is no longer a prime factor. You move on to 2+1 = 3 15%3 = 0. 3 is a prime factor. 15/3 = 5 5%3 != 0. So 3 i not a prime factor more than once. You move on to 3+1 = 4 (You can skip even numbers here as well) 5%4 != 0. So move on to 4+1 = 5 5%5 = 0. S0 5 is a prime factor. 5/5 = 1. So the loop ends with 2,2,3,5 as the prime factors. This is the method we used in school to find prime factors.

Wow. It is beautiful and simple but I never thought like that. But please now that it's not in school anymore, do you have a much better approach?

Why is this not efficient? It is much more efficient than the code you wrote originally

And after it outputs the prime factors, how do you find the highest factor?

The number which sets n to 1 will be the highest factor. In the example that I gave above 5 will be the highest factor

Implement probabilistic algorithms...there is no trick here. Prime Factorization (a simplified version of integer factorization) is an area of research. If you can implement an efficient algorithm then cryptography as we know it today will be at risk.

Implement probabilistic algorithms...there is no trick here. Prime Factorization (a simplified version of integer factorization) is an area of research. If you can implement an efficient algorithm then cryptography as we know it today will be at risk.