Array of size n

[marcnadeau97]

Hello,

I've started using C++ just this week. I've been using matlab for a few years. I want to do an N size array and I get this error with visual studio. Anyone can help me with that?

  #include <iostream>
#include <cmath>
using namespace std;


int main()
{
    double limit = 8;
    int step = 100;
    double w = limit / step;
    double x[step];
    x[0] = 0;

    for (int i = 1; i < step; i++) {
        x[i] = x[i - 1] + w;
    }
    system("pause>0");
}

error C2131: expression did not evaluate to a constant
message : failure was caused by a read of a variable outside its lifetime
message : see usage of 'step'

[mikef]

Make int step a const.

const int step = 100;

[thmm]

Don't learn old stuff. Prefer std::array over C-style arrays.
https://isocpp.github.io/CppCoreGuidelines/CppCoreGuidelines#Res-stack

[marcnadeau97]

well that was easy

[marcnadeau97]

Can someone explain me how I can allocate more memory to avoid stack overflow errors? I need to deal with big array but I always get stack overflow error. again using visual studio.

#include <iostream>
#include <cmath>
#include <array>
using namespace std;


int main()
{
    double limit = 8;
    const int step = 1000000;
    double w = limit / step;
    std::array<double,step> x={0};

    for (int i = 1; i < step; i++) {
        x[i] = x[i - 1] + w;
    }
    cout << "complete";
    return 0;
}

[thmm]

For large arrays you can use std::vector.

[marcnadeau97]

What am I doing wrong now? ahah

it just ccrashes

#include <iostream>
#include <cmath>
#include <vector>
using namespace std;


int main()
{
    double limit = 8;
    int step = 1000;
    double w = limit / step;
    std:: vector<double> x={0};
    for (int i = 1; i < step; i++) {
        x[i] = x[i - 1] + w;
    }
    cout << "complete";
    return 0;
}

[jonnin]

What am I doing wrong now? ahah

it just ccrashes

you didn't look up vector or study it, is the harsh answer.
a quick google might have taken you here: https://www.geeksforgeeks.org/default-value-of-vector-in-c-stl/

your vector is empty and there you are accessing it. you can use push back or preallocate to give it some space.
std:: vector<double> x(step,0); //size, value
for (int i = 1; i < x.size(); i++)

it looks a lot like x[i] = i*w;
https://www.cplusplus.com/reference/algorithm/generate/ is overkill, but since you are learning about what is available... take a look!

[marcnadeau97]

oh yeah that's right. I've been playing around too much and I confuse the 2.

[seeplus]

If you know the size of the array at compile time, then you can use std::array or even just a c-style array. If you get stack overflow error, then mark the array definition as static - or define as a global (before main).

#include <iostream>
using namespace std;

int main() {
	const int step = 100;
	const double limit = 8;
	const double w = limit / step;
	static double x[step];

	x[0] = 0;

	for (int i = 1; i < step; i++) {
		x[i] = x[i - 1] + w;
	}

}

Just for info. Note that in c/c++ int division gives an integer! So 5 / 7 gives 0 - even if the result is assigned to a type double! If you want a non-integer result, then at least one of the divisor and dividend needs to be float/double.

[jonnin]

I've been using matlab for a few years

I missed that note before. Its fair to say that matlab and C++ approach how to code from polar opposite positions. Matlab was designed to be easy to use by non-coders, math people that just want to express a problem quickly and easily. C++ was designed for professional coders who need to know every detail, matlab HIDES a ton of details (type, dimensions, hidden copying, more (modern c++ is bad in a few places about the hidden loops and hidden copies as well) ). Matlab seems to have a loose C syntax root, so at times things will feel familiar, but even so be prepared for some shocks along the way.