write a function to round off numbers

Write your own roundn function that takes a floating point value as its input and rounds its nth decimal digit to the nearest number.

>> res = roundn(5.672,3); % results in res = 5.67
>> res = roundn(5.677,3); % results in res = 5.68
>> res = roundn(-5.672,3); % results in res = -5.67
>> res = roundn(-5.677,3); % results in res = -5.68

Answer 1

[Image Analyst]

Try multiplying by a factor of 10 then rounding and dividing. That should be a good hint - probably too much of one.

[Zcarfaz]

ok I got this for but when i run the function it won't give the right answer but when I do it manually on the command window it works. also this is not the way i have to do it I need to write a for loop to check the value of the myfloat and then change the factor of 10 accordingly. please help. thanks!

function output = roundn(myinput,myfloat)

      if myfloat == 1

          output = round(myinput*10)/10;
          fprintf('%.1f\n', output);
      end

      if myfloat == 2
          output = round(myinput*100)/100;
          fprintf('%.2f\n', output);
      end

      if myfloat == 3
          output = round(myinput*1000)/1000;
          fprintf('%.3f\n', output);
      end
  end

[Image Analyst]

Instead of checking for every possible factor of 10, why don't you just use 10^myFloat?

And how are you running it, if not from the command line? What value are you passing in when it does not run?

Answer 2

[Walter Roberson]

This cannot be done in MATLAB, not if the values are to be stored numerically as datatype single() or double().

>> fprintf('%.999g\n', 0.1)
0.1000000000000000055511151231257827021181583404541015625

This establishes that even if you had 0.1 already given, that you do not actually get exactly 0.1 stored.