UsingWW: Numerical tolerance for very small numbers

WeBWorK Problems

Numerical tolerance for very small numbers

This forum has a limit to the number of forum postings you can make in a given time period - this is currently set at 10 posting(s) in 1 day

Hi, I was writing this simple code snippet:

$a = 0.6**20;

Enter a: [_______]{$a}

When I input 0.6**20 as the answer, it is incorrect. I suspect this has to do with the tolerance issue. So I add

Context('Numerics')

Context()->flags->set(tolerance=>0.0001);

The answer still is not correct.

Could you help with this problem? Thanks.

Re: Numerical tolerance for very small numbers

by Davide Cervone - Thursday, 1 November 2012, 5:36 PM

Try

    $a = Real(0.6**20);

as this will make $a a MathObject, which will be handled better. As a Perl real, $a is converted to a string and then passed to Compute() to be converted to a MathObject, but the string version of this Perl real is 3.65616e-5 which MathObjects considers to be 3.65616 times e minus 5 rather than scientific notation (which would be 3.65616E-5).

Alternatively, you could use

    Enter a: [_________]{uc($a)}

to force the lower-case "e" to an upper-case one.

Davide