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
I find it's much more foolproof to store work in units of kT whenever we need to store them as a unitless array. We appear to use kJ/mol in the code at various places, such as here, which has the potential to cause some confusion at best and errors at worst (if we accidentally feed them to, say, EXP or BAR before converting to kT).
If we do store energies or work with specific units, we should make sure to actually store unit-bearing quantities. Code like this has the potential to cause confusion by returning unitless quantities that must be interpreted in kJ/mol.
The best policy is
If the quantity is supposed to be unit-bearing, make sure it is a simtk.unit.Quantity with the appropriate units
If you want it to be dimensionless, convert it to units of kT; it's then safe to be unitless
The text was updated successfully, but these errors were encountered:
jchodera
changed the title
Store energies in units of kT throughout
Potential unit inconsistencies
Apr 16, 2017
I find it's much more foolproof to store work in units of kT whenever we need to store them as a unitless array. We appear to use kJ/mol in the code at various places, such as here, which has the potential to cause some confusion at best and errors at worst (if we accidentally feed them to, say, EXP or BAR before converting to kT).
If we do store energies or work with specific units, we should make sure to actually store unit-bearing quantities. Code like this has the potential to cause confusion by returning unitless quantities that must be interpreted in kJ/mol.
The best policy is
simtk.unit.Quantity
with the appropriate unitsThe text was updated successfully, but these errors were encountered: