From: fateman@cs.berkeley.edu   
      
   On 12/13/2017 8:50 PM, Albert Rich wrote:   
      
   >   
   > No, if all the application conditions of any of Rubi's 6000+ integration   
   rules is satisfied, the result of applying that rule will be a mathematically   
   valid antiderivative.   
   >   
   > If the simplification and expansion functions defined in Sympy's utility   
   file for Rubi do not exactly mimic Mathematica's, the resulting    
   ntiderivatives may not be quite as concise as those produced by Rubi running   
   on Mathematica, but they still will    
   terminate and be mathematically correct.   
      
   I suspect that the issue I encountered was that intermediate results   
   from some of the rules (part-way toward the final result) were in   
   unanticipated  (less concise?) forms than expected, and this inhibited   
   the appropriate application of additional rules.   
   Whether something like this happens in sympy, or whether this suspicion   
   is just wrong, I don't know.   
   Translating the rules to if-then-else programs would probably eliminate   
   some of the problems related to matching per se, and maybe solve some of the   
   sympy issues.   
      
   RJF   
      
      
      
      
   >   
   > Albert   
   >   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)