Registered Member
|
Using the minimal test program below, with both Eigen 3.3 and the latest Hg head (as of today), I receive the following compilation error with GCC 6.3 (on Linux):
The exact same code compiles and runs fine if "long double" is replaced with "double". Any idea why using long doubles here causes a compilation error? I'm not familiar enough with the Eigen codebase to make much progress following the errors. Thanks Minimal test:
|
Registered Member
|
Forgive my mistake - I continued to use the regular "double" vector typedefs, which is the cause of my problems. Using the proper long double matrix types for the vectors f and u seems to allow the code to compile and run.
Thanks |
Registered users: abc72656, Bing [Bot], daret, Google [Bot], Sogou [Bot], Yahoo [Bot]