20 April, 2016

More on bugs in Mathematica

In comments to my previous post, Rytis Jursenas pointed at an entire article in the Notices of AMS discussing several bugs in Mathematica. Since such problems are apparently quite common, I feel it deserves a separate post.

Here it comes:


(just in case, here is the preprint)


"...We have been using Mathematica as a tool in our mathematical research. All our computations with Mathematica have been symbolic, involving only integers (large integers, about 10 thousand digits long) and polynomials (with degree 60 at most), so no numerical rounding or instability can arise in them, and we completely trusted the results generated by Mathematica. However, we have obtained completely erroneous results."
"...Software bugs should not prevent us from continuing this mutually beneficial 
relationship in the future. However, for the time being, when dealing with a problem whose answer cannot be easily verified without a computer, it is highly advisable to perform the computations with at least two computer algebra systems."

Well, I didn't expect such a conclusion to be drawn in 2014.

Take care,

Misha

2 comments:

Brittany Banks, Wolfram Research said...

Thanks for being a fan of our products. The bug mentioned here was fixed over a year ago, in open communication with the users who reported it. For more details, please see the response from Daniel Lichtblau in the Wolfram Community: http://community.wolfram.com/groups/-/m/t/374658?p_p_auth=qBppEnx7. The issue with ClebschGordan returning a symbolic result that is invalid for some values of the symbolic parameter is one we are aware of. It is a limitation in the restrictions we use in the symbolic form of the function. We will consider extending this to give a complete set of conditions for validity after numeric substitution.

Lemeshko said...

Dear Brittany,

Many thanks for your detailed response!

Misha