MakeCode Arcade Precision Activities - Solution
The following MakeCode Arcade programs can help you see approximation and overflow errors in action.
Floating Point and Overflow
Click on Edit Code to open the program.
When you run this program, it is dividing 17 by 5 and displaying the result.
- Replace the 17 with 3.4. What is the calculation now? Answer: 0.67999999999999999
- Perform the same calculation on a calculator. What is the answer? Answer: 0.68
- Why do you think they are different? Answer: Different precision levels
- Replace 3.4 with it with 0.00000000001. What is the calculation now? Answer: 2e-12 the numbers after the ‘e’ are an exponent of 10
- Now, change the calculation to a multiplication rather than division.
Once you change to multiplication change the number to calculate a large number. Keep changing the numbers until it reaches overflow.
- Now try using the myint and myfloat variables in your calculation. Hint - look in the Variables Toolbox drawer. Answer: Floating point 534.96
Is the result an integer or a floating point number?
- From the Info Toolbox drawer, pull out a Set Score block and place after the mySprite Say block. From the Text Toolbox drawer, pull out a Parse to Number block and place in the Set Score block. In the Parse to Number block type "7.5e2".
What happpens? Answer: Displays a score of 750 because 7.5 x 10^2 = 750 MakeCode Arcade Parse Float
Approximation
Click on Edit Code to open the program.
What is the value of x? Answer: 0.8
What is the value of y? Answer: 0.8
According to the algorithm, is the program correct, when it displays "False"? Answer: No - these are equivalent values so it should return True
Why or why not? What is going on? Hint: move the mySprite say block
out of the source code and replace it with the other disabled mySprite say block on the workspace
Use the drop-down menu to change the variables. Answer: The X variable value shows a 0.79999999 instead of 0.8. This is an example of a precision rounding error.