The following modules are too narrow in scope to exist as standalone modules:
partitions
bernoulli
At the same time, there are too many special-purpose functions in the arb module.
Idea: maybe create a dedicated module for combinatorial/number-theoretic functions. This could host partitions, bernoulli and functions like bell, fib currently located in the arb module. What would be a good name for such a module?
arb_combinat? (...but the functions are not strictly combinatorial)
arb_arith / arb_arithfun? (...but the functions are not strictly arithmetical)
arb_sequences? (...but some functions don't naturally correspond to linear sequences)
arb_integers? (...but some functions return rational numbers)
It's a bit annoying to give all functions prefixes based on the module, but maybe this convention could be relaxed. We might want to do this elsewhere, e.g. provide arb_erf as an alias for arb_hypgeom_erf.
Related:
hypgeom could just be merged with arb_hypgeom.
There have also been complaints about about dirichlet.h being a too generic name.
The following modules are too narrow in scope to exist as standalone modules:
partitionsbernoulliAt the same time, there are too many special-purpose functions in the arb module.
Idea: maybe create a dedicated module for combinatorial/number-theoretic functions. This could host
partitions,bernoulliand functions likebell,fibcurrently located in thearbmodule. What would be a good name for such a module?arb_combinat? (...but the functions are not strictly combinatorial)arb_arith/arb_arithfun? (...but the functions are not strictly arithmetical)arb_sequences? (...but some functions don't naturally correspond to linear sequences)arb_integers? (...but some functions return rational numbers)It's a bit annoying to give all functions prefixes based on the module, but maybe this convention could be relaxed. We might want to do this elsewhere, e.g. provide
arb_erfas an alias forarb_hypgeom_erf.Related:
hypgeomcould just be merged witharb_hypgeom.dirichlet.hbeing a too generic name.