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
according to the code, currently, tolerance is specified in the units of #contacts.
An attempt to trace units: line 135: margs are in the units of #contacts (i.e. proportional to # of reads in the library) line 145: nzmargs are in the units of margs (#contacts) line 156: var in the units of nzmargs (#contacts) line 167: tol is compared to var
Specifying tol in relative units makes more sense since it is hard for the user to know the # of reads in the library before balancing. We can still keep the option to specify an absolute tolerance, but I would argue that relative tolerance is a better default.
The text was updated successfully, but these errors were encountered:
according to the code, currently, tolerance is specified in the units of #contacts.
An attempt to trace units:
line 135: margs are in the units of #contacts (i.e. proportional to # of reads in the library)
line 145: nzmargs are in the units of margs (#contacts)
line 156: var in the units of nzmargs (#contacts)
line 167: tol is compared to var
importantly, margs are re-computed using biases, but biases average out to 1.0, so margs first come out in the units of #contacts
Specifying tol in relative units makes more sense since it is hard for the user to know the # of reads in the library before balancing. We can still keep the option to specify an absolute tolerance, but I would argue that relative tolerance is a better default.
The text was updated successfully, but these errors were encountered: