C++ Using variables in equations; error: expression must have integral or unscoped enum type & others

Question

srand( 0 ); int points; // number of points float computerNumber; // number generated by the computer float guess; // user's guess char quit; // What the user enters when they want to quit int totalPoints; //the total score of all of the games played int avgPoints; // the average score of all games played int gamesPlayed; // how many games have been played float rangeLow; // the lower end of the range float rangeHigh; // the higher end of the range points = 5; quit = 'n'; gamesPlayed = 0; totalPoints = 0; while ( quit != 'q' ) { gamesPlayed++; cout << "Welcome to Guessing Game! \n"; points = 5; cout << "What would you like your range to be? \n"; cout << "Low number: \n"; cin >> rangeLow; cout << "High number: \n"; cin >> rangeHigh; if ( rangeLow > rangeHigh ) { cout << "Please use a high number that is greater than the low number. \n"; cout << "Low number: \n"; cin >> rangeLow; cout << "High number: \n"; cin >> rangeHigh; } else { ; } computerNumber = rand( ) % (rangeLow - rangeHigh + 1) + 10; cout << "Computer Number: " << computerNumber << endl; cout << "Points:" << points << endl; cout << "what is your guess? \n" << endl; cin >> guess; cout << "Your guess is: " << guess << endl;

When I input this code (among other error-free lines of code that don't affect these lines), it won't compile and outputs two error messages- "expression must have integral or unscoped enum type" and "'%' is illegal, right operand has type 'float'"

I have a feeling that it has something to do with using variables in my equation, but that shouldn't be a problem? All of the variable types are float that have to do with this equation, and I'm pretty confused.

There's fmod but I don't think OP wants to use float here at all. — 001
– 001, Commented Oct 23, 2019 at 18:22

Ted Lyngmo · Accepted Answer · 2019-10-23 19:28:31Z

% is the modulus operator for integer values. Your range endpoints are floating point variables. In your exercise, you should make the endpoints, computerNumber and guess into int variables.

int rangeLow, rangeHigh, computerNumber, guess;

Then take a look at the computer's number generator and insert some values to try it out:

computerNumber = rand() % (rangeLow - rangeHigh + 1) + 10; rand() % ( 5 - 20 + 1) + 10; rand() % -14 + 10;

So, you'll have rand() % -14 + 10. That doesn't look correct. Reverse the range endpoints:

computerNumber = rand() % (rangeHigh - rangeLow + 1) + 10; rand() % ( 20 - 5 + 1) + 10; rand() % 16 + 10;

This, rand() % 16 + 10, is better, but will obviously create a number in the range [10, 10+16), or to put it differently (as a closed interval), [10, 25]. If you replace +10 with +rangeLow like this:

computerNumber = rand() % (rangeHigh - rangeLow + 1) + rangeLow;

You now get the range like this:

[rangeLow, rangeLow + rangeHigh - rangeLow + 1) => [rangeLow, rangeHigh + 1) (left-closed interval) => [rangeLow, rangeHigh] (closed interval)

Which means [5, 20] with the example values I used.

Now, take a look at <random>. That part of the standard library contains classes and functions to make all this much simpler:

FutureJJ · Accepted Answer · 2019-10-23 18:53:04Z

So, the error is % operator cannot be used with float values, it can only be used with integer(int) data types.

You need to typecast float to an int, one method is by using static_cast<int>(yourFloatNumber) so that your calculation code line looks like this:

computerNumber = rand( ) % (static_cast<int>(rangeLow) - static_cast<int>(rangeHigh + 1)) + 10;

The modulus operator (%) has a stricter requirement in that its operands must be of integral type.

Another solution enlightened by @NathanOliver in the comments: If you are ok with using int instead of float for rangeLow and rangeHigh all across your code then typecasting wouldn't be required at all.

Reference

@NathanOliver An elaborate explanation at stackoverflow.com/a/26269263/7040601
I'm not asking why use static cast vs c-style cast, I'm asking why cast at all. Since you are only using them in an integer context, if you just change them to int's instead of floats's then you don't need all that verbosity.
It would be good only if OP wants the variable in problematic line to be int, if he wants to accept input as float and do some float operations then int won't work.

Collectives™ on Stack Overflow

C++ Using variables in equations; error: expression must have integral or unscoped enum type & others

2 Answers 2

Comments

4 Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

4 Comments

Linked

Related