Deep recursion on subroutine with bignum and Astro::Units

[Anonymous Monk]

I'm doing some stuff with the module Astro::Units whose pod suggests to explicitly use bignum. The program works fine until I try doing some simple arithmetic. I can't reproduce the error with an SSCE so here are the errors and the calculations causing them:

Deep recursion on subroutine "Math::BigFloat::bdiv" at /5.26.2/Math/BigInt.pm line 1867.
Deep recursion on subroutine "Math::BigInt::Calc::_lsft" at /5.26.2/Math/BigFloat.pm line 1948.
Deep recursion on anonymous subroutine at /5.26.2/Math/BigInt/Calc.pm line 1412.
Deep recursion on subroutine "Math::BigInt::bdiv" at /5.26.2/Math/BigInt.pm line 54.

Here's the math and the results when loading and not loading (but not directly using) both bignum and Astro::Units:

perl -le '$Da=406700; $Dp=356400; $Dm=361918; $pct = ($Da - $Dm) / ($D
+a - $Dp); print $pct'

0.890298210735586

perl -Mbignum -le '$Da=406700; $Dp=356400; $Dm=361918; $pct = ($Da - $
+Dm) / ($Da - $Dp); print $pct'

0.8902982107355864811133200795228628230616

perl -Mbignum -MAstro::Units -le '$Da=406700; $Dp=356400; $Dm=361918; 
+$pct = ($Da - $Dm) / ($Da - $Dp); print $pct'

0.8902982

perl -MAstro::Units -le '$Da=406700; $Dp=356400; $Dm=361918; $pct = ($
+Da - $Dm) / ($Da - $Dp); print $pct'

0.890298210735586

Can anyone see what's going on here? That third result is puzzling. What could be causing such errors? Thanks!

[hv]

On a general note, I consider it unwise of the module author to use bignum and doubly-unwise for them to recommend that everyone using their module should do so. Each of bignum, bigint, bigrat can be fine and useful for simple one-liners, but for anything more complex they are very likely to cause problems.

For this specific issue, note that the current Math::BigInt maintainer has made wide-reaching changes multiple times in recent months in an attempt to make the interactions between the different classes behave more sanely in the context of use bignum; it is not clear to me whether it has now reached a stable point. So you may want to try both more recent and older versions of the distribution - check the changelog for references to "upgrading and downgrading".

For getting the calculations you are using to work reliably, I suggest ditching bignum; instead instantiate your numbers as Math::BigFloat, and hope that the use of bignum inside Astro::Units doesn't result in unexpected conversions to Math::BigInt or Math::BigRat.

Note that you should also be able to change the Math::BigFloat configuration after loading Astro::Units, if the latter's choices are not suitable for your needs.

[Anonymous Monk]

Thank you Monks! I'm still struggling to isolate the problem to an SSCE but did make an interesting discovery. Turns out the math is not causing the error. I though it was because commenting that line out eliminated it. Just before doing the math I use "int" on one of the numbers. If I don't use "int" the program works! It also works if "int" is used and "use bignum" is commented out AND this is done: Math::BigFloat->config({ div_scale => 1 }); (don't need to use Math::BigFloat, cause I guess Astro::Units already loaded it). I have no idea what is going on but will just not use "int". Hope this makes sense. Thanks again!

[soonix]

I added an issue to https://github.com/spajai/Astro-Units/issues/1 assuming that this is the original developer's repo.

[karlgoethebier]

I always thought it was like converting miles to kilometers. You don't really need a Perl module for that. Does not make a big difference whether AU or nautical miles, light years or whatever - except the size of the numbers. But I am only a hobby astronomer and anyway very bad in math.

[Anonymous Monk]

I always thought it was like converting miles to kilometers. You don't really need a Perl module for that.

Thanks for the reality check. Converting AU to KM is easier than using the module. Multiply by 149597870.7 and done.

[Anonymous Monk]

See bug report, this Astro::Units won't even install with modern Math::BigInt. I don't know what's the purpose of line 12, and why setting div_scale to 0 with older Math::BigInt results in 7 digits you observe in your third case, but that's direct consequence.

I'd suggest installing latest Math::BigInt, fixing line#12 (change value to 1? don't know) and see if deep recursion problem goes away.

[Anonymous Monk]

OP here. Forgot to mention after the errors Perl uses a lot of CPU for about 10 seconds and dies with: Segmentation fault: 11

[Anonymous Monk]

OP again. Perl 5.26.2 with module versions:
• Astro::Units 1.10 (No other versions)
• bignum 0.47
• Math::BigInt 1.999816