This PR addresses issue #13. (Done in collaboration with @hrnasif )
For added flexibility, a pair of module-level global variables that describe the floating point dtype for TF and NP were created by adding __inti__.py . Then, every function that specifies dtype as an input was altered to take the new global variables as the new input, allowing the user to switch between float32 and float64 by changing the values in __inti__.py.
Of note: this PR was done based on the 2-libsbn-ratiobranch, not from master
This PR addresses issue #13. (Done in collaboration with @hrnasif )
For added flexibility, a pair of module-level global variables that describe the floating point dtype for TF and NP were created by adding
__inti__.py. Then, every function that specifiesdtypeas an input was altered to take the new global variables as the new input, allowing the user to switch betweenfloat32andfloat64by changing the values in__inti__.py.Of note: this PR was done based on the
2-libsbn-ratiobranch, not frommaster