You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is a great library for interval computation and optimization ! But I am having a problem and hope the author can help.
I build the library following the direction in section "For Windows Users (Win64 version)" at this link, with Microsoft Visual Studio 14.0 Win64. All seems OK.
But then I test the example "doc-arithmetic.cpp", there is an error
when I start the program. Calling the stack shows the position is in line /** * construct rounding control **/ template<bool C> inline void filib::rounding_control<double,C>::setup() { /** * reset * double precision, round to nearest, no * interupts **/ _controlfp( _EM_INVALID|_EM_DENORMAL|_EM_ZERODIVIDE|_EM_OVERFLOW|_EM_UNDERFLOW|_EM_INEXACT ,MCW_EM); _controlfp(_PC_53,MCW_PC); _controlfp(RC_NEAR,MCW_RC); }
of file "rounding_control_double_asmmsvci386.icc". Then I press the button "Retry" and "Continue", the program goes on. The results seems to be fine except the PI value, which in the result file gives: PI = [8.6192e+97, 8.61921e+97] 2 PI = [1.72384e+98, 1.72384e+98] 1/2 PI = [4.3096e+97, 4.3096e+97]
So my question is, if this library can only be built and used as 32 bit ? Or the problem results from the filib++ and I should use another interval library ?
Looking forward for any reply ! Thanks !
The text was updated successfully, but these errors were encountered:
Thanks for your prompt reply ! The error didn't occur with this pre-compiled version. But if I compile the source code here by myself, the error remains. I will use this pre-compiled version temporarily to test my model and compile by myself later if necessary. Thanks again for your time !
This is a great library for interval computation and optimization ! But I am having a problem and hope the author can help.
I build the library following the direction in section "For Windows Users (Win64 version)" at this link, with Microsoft Visual Studio 14.0 Win64. All seems OK.
But then I test the example "doc-arithmetic.cpp", there is an error
when I start the program. Calling the stack shows the position is in line
/** * construct rounding control **/ template<bool C> inline void filib::rounding_control<double,C>::setup() { /** * reset * double precision, round to nearest, no * interupts **/ _controlfp( _EM_INVALID|_EM_DENORMAL|_EM_ZERODIVIDE|_EM_OVERFLOW|_EM_UNDERFLOW|_EM_INEXACT ,MCW_EM); _controlfp(_PC_53,MCW_PC); _controlfp(RC_NEAR,MCW_RC); }
of file "rounding_control_double_asmmsvci386.icc". Then I press the button "Retry" and "Continue", the program goes on. The results seems to be fine except the PI value, which in the result file gives:
PI = [8.6192e+97, 8.61921e+97] 2 PI = [1.72384e+98, 1.72384e+98] 1/2 PI = [4.3096e+97, 4.3096e+97]
So my question is, if this library can only be built and used as 32 bit ? Or the problem results from the filib++ and I should use another interval library ?
Looking forward for any reply ! Thanks !
The text was updated successfully, but these errors were encountered: