1. Check out our app, Forums for Android! Download and leave feedback here!

Calculator Error

Discussion in 'Android Devices' started by Eric61323, Mar 3, 2010.

  1. Eric61323

    Eric61323 Member
    Thread Starter
    5

    Mar 3, 2010
    6
    0
    5
    Machinery Sales Rep
    Maryland
    The calculator app that comes with the Android OS apparently has an error. I've confirmed this error with two other Moto Droids and one HTC Android phone.

    In the calculator app enter:

    3322.72 - 3321.92 then = and the answer is .799999

    The correct answer is .80

    This may seem like a minor point (.000001 error) but the answer is still wrong! Basic arithmetic operations should be a no brainer...

    A Verizon service rep confirmed this error on the Droid and Eris phones he had in their call center. I wonder if 2.1 will fix this?

    Does anyone else get this result?
     

    Advertisement

  2. gonzoguy24

    gonzoguy24 Well-Known Member
    118

    Nov 11, 2009
    1,104
    41
    118
    Structural Engineer - Nuclear Industry
    Aurora, IL
    That is hilarious...something is going on with the computational method used by the calculator...I would not worry about it...unless you are in 1st grade and cheating on your arithmetic test :)
     
  3. Les Common

    Les Common Well-Known Member
    36

    Oct 29, 2009
    61
    5
    36
    cue the Beach Boys" "Round round round round, I cannot round..."


    It isn't the Droid, it's the Droid stock gonkulator.

    RealCalc gives a nice clean 0.8.

    -Les
     
    AndremG likes this.
  4. PridgNYC

    PridgNYC Well-Known Member
    18

    Nov 16, 2009
    99
    13
    18
    Yep, definately a bug in the built in calc app, RealCalc (free on the marketplace) works fine though.
     
  5. Fabolous

    Fabolous Superuser
    313

    Nov 7, 2009
    3,635
    2,126
    313
    Multi-platinum Rapper
    Illinois
    Haha, this is a funny bug. I can confirm it still exists in 2.1, at least in the ERE25 build.
     
  6. Trapper31

    Trapper31 Well-Known Member
    36

    Dec 9, 2009
    109
    6
    36
    Cleveland
    I was about to suggest RealCalc as well but you beat me to it! I tried it out and it works fine.
     
  7. obutto

    obutto Well-Known Member
    38

    Nov 27, 2009
    307
    14
    38
    I don't know if your joking or your serious....TI-89 tends to do this type of round as well...I don't know why you would need to be that accurate..
     
  8. coloradogregg

    coloradogregg Well-Known Member
    16

    Feb 26, 2010
    62
    1
    16
    Construction
    Dallas, TX
    How did you even find this out? You must have more free time than I do.
     
  9. Jim Dawson

    Jim Dawson Well-Known Member
    36

    Nov 27, 2009
    194
    9
    36
    East Alton, IL
    I think he tried every numeric combination he could think of until he found one that didn't work.:)

    How long did that take?
     
  10. Keyser-Soze

    Keyser-Soze Well-Known Member
    18

    Jan 8, 2010
    71
    10
    18
    mathematically, .79999 repeating =.8, if the 9 repeats forever.

    to see this, write .79 repeating equals x.

    10x =7.9 repeating.
    10x-x =7.2 = 9 x
    x=.8
    there are other proofs. it's well established

    perhaps the droid just appreciates the subtlety of math :)
     
    johnlgalt likes this.
  11. Keyser-Soze

    Keyser-Soze Well-Known Member
    18

    Jan 8, 2010
    71
    10
    18
    gr double post
     
  12. esocid

    esocid Well-Known Member
    56

    Dec 8, 2009
    122
    4
    56
    Not worth mentioning
    Virginia Beach, VA
    1 + 1 = 3, for large values of 1.

    Pretty funny that the 3 people whose job it was to program a calculator in android missed this, albeit small, error.
     
  13. messenger13

    messenger13 Well-Known Member
    163

    Dec 8, 2009
    3,133
    381
    163
    Director of IT
    Cleveland, OH
    Yep, I tested RealCalc as well. I wasn't surprised that it was correct.
     
  14. CegAbq

    CegAbq Well-Known Member
    38

    Nov 15, 2009
    109
    12
    38
    Albuquerque
    Further strange results:
    1.72 -.92 = 0.8
    22.72 -2 1.92 = 0.8
    322.72 - 321.92 = 0.8000000
     
  15. empyreandance

    empyreandance Well-Known Member
    16

    Feb 18, 2010
    81
    2
    16
    your proof is wrong... by what you set x to be, 10x-x=7.19999... assuming it's 7.2 is assuming what you're trying to prove

    the correct proof would be to take the limit of its p-adic expansion.
     
  16. Keyser-Soze

    Keyser-Soze Well-Known Member
    18

    Jan 8, 2010
    71
    10
    18
    Oh no! another product of poor math education. Or perhaps you're using the Droid calculator?

    7.999999999999999...
    - .799999999999999...
    =7.2

    No ambiguities bub.

    Consider .9999999999...=x
    10x=9.99999999...
    x=.99999999...
    9x=9 x=1
     
  17. empyreandance

    empyreandance Well-Known Member
    16

    Feb 18, 2010
    81
    2
    16
    No, a product of a math degree. First of all, don't appreciate the condescending attitude. Second, I would be more than happy to show you the fault in your logic.
     
  18. Eric61323

    Eric61323 Member
    Thread Starter
    5

    Mar 3, 2010
    6
    0
    5
    Machinery Sales Rep
    Maryland
    Yeah, I thought this was pretty funny as well - it's definitely a programming problem as the error occurs across different OS versions and hardware platforms.

    I came across this error by chance while balancing my check book and when I came up with more than two digits after the decimal point I knew something wasn't right...

    Because I'm not calculating the trajectory of a rocket launch to the moon I'm not going to worry about such a small error. Even so, I'll probably get Realcalc...
     
  19. Keyser-Soze

    Keyser-Soze Well-Known Member
    18

    Jan 8, 2010
    71
    10
    18
    I double dog dare you to prove me wrong. But by PM, so this doesn't clog the thread. though you should come back here to admit it when you realize you're wrong.
     
  20. ubergray

    ubergray New Member
    6

    Jan 14, 2010
    4
    1
    6
    Not quite a calculator error but a throwback from floating point arithmetic. This is particularly common with computations in programming, e.g., you do not test for equality to zero but rather test for proximity to machine-epsilon: abs(x)<eps.

    Also, the comment above about .799999... = .8 is correct.
     
    johnlgalt likes this.
  21. doomedromance

    doomedromance Well-Known Member
    78

    Nov 9, 2009
    664
    82
    78
    The Left Coast
    I tested this with a piece of paper and a R.S.V.P fine point pen and I get 0000.80
    When I can go to the store tomorrow and buy a pencil I'll run the test again and see what my result is.
     
    johnlgalt likes this.
  22. psteichen

    psteichen Member
    5

    Jun 13, 2010
    5
    0
    5
    I remember from my advanced math education (in middle school) that .799999999...=.8, although I've never heard of epsilon. But I think everyone can agree that writing it as such is bad form, and this is clearly a glitch that should be fixed. Perhaps they should hire someone smart like ubergray to fix it.

    It confused me for a while when I saw that 601.1-606.3=-5.199999

    To whom should one address such a bug fix request?
     
  23. AndremG

    AndremG Well-Known Member
    36

    May 15, 2010
    59
    4
    36
    Nashville, TN
    Les is right. The OP has an issue with significant figures. Since the less significant place in either of the numbers is 0.0x, any result should be rounded to that place. Of course, the really fault does lie in the calculator, but the operator has to know not to blindly use what the calculator gives him.
     
  24. talkingtoes

    talkingtoes New Member
    5

    Jul 27, 2010
    1
    0
    5
    Windows 3.11 had this bug too. (if you can find a copy)

    3.11 - 3.1 is not 0.01 on the phone either.
     
  25. smacky

    163

    Jan 11, 2010
    2,844
    299
    163
    Get on the phone, tell them this, demand an upgrade to Droid X.
     

Share This Page

Loading...