home bbs files messages ]

Forums before death by AOL, social media and spammers... "We can't have nice things"

   sci.math.symbolic      Symbolic algebra discussion      10,432 messages   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]

   Message 8,800 of 10,432   
   Nasser M. Abbasi to Albert Rich   
   Re: Rubi 4.8 released   
   02 Jun 15 22:51:27   
   
   From: nma@12000.org   
      
   On 6/2/2015 5:12 PM, Albert Rich wrote:   
      
   > If the optimal antiderivative involves the function Int, no closed-form   
   >antiderivative exists (unless there is an error in the Rubi test suite).   
   >Therefore, I recommend Nasser's test program check to see if the optimal   
   antiderivative   
   >involves Int, and if so judge antiderivatives involving the integration   
   function a success.   
   >   
      
   The above will raise the score of all 3 CAS's by the   
   same amount, since I am almost sure now if the optimal is   
   not known by Rubi, the other CAS'es did not solve it (but will   
   check on this, please see below).   
      
   So instead of the current score of  (Rubi, Mathematica, Maple)   
   95.5%, 92.2%, 83.8%   
   will become (adding say the 4.5% for those optimal results with Int)   
   99%, 96.7%, 88.3%   
      
   So it will just shift all of them by same amount.   
      
   But it seems to me it would be confusing to call a test success if   
   a CAS did not actually solve the integral? Even if there is no known   
   closed form antiderivative.   
      
   So I think it might be better to keep this part as is. Also, the   
   optimal result is given in the result with the CAS result in the   
   same section, so one can see that there is no optimal result for   
   the failed integral.   
      
   > Of course, if the test program encounters a problem whose optimal   
   antiderivative   
   >involves Int AND one of the systems returns a valid closed-form   
   antiderivative,   
   >that problem should be flagged so I can correct the optimal antiderivative in   
   the test suite.   
   >   
   > Albert   
   >   
      
   This is great idea. I will add a post-processing step to check for this.   
      
   Any CAS which solves a problem that has `Int` in Rubi's optimal field,   
   will go to a special list, and will add a new check field the tables in   
   the final result to make it easier to spot these problems.   
      
   The tests take long time to run, about one week or so. So this   
   will take sometime to be done.   
      
   regards,   
   --Nasser   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]


(c) 1994,  bbs@darkrealms.ca