C++20 concepts that constrain the value and range types accepted by Statistical analysis tools.

The statistics module is generic in its sample type: it accepts plain scalars as well as nda arrays, but only for a restricted set of element types. The three concepts in this group spell out those restrictions:

triqs::stat::StatCompatible — a single sample value. Either a double / std::complex<double> scalar, or an nda::Array whose value type is double / std::complex<double> and whose algebra is 'A' (i.e. it behaves as a generic array under elementwise operations).
triqs::stat::AccCompatible — additionally required by the streaming accumulators (triqs::stat::lin_binning, triqs::stat::log_binning). It refines triqs::stat::StatCompatible by also requiring scalars or regular array types, so that bins can own their storage.
triqs::stat::StatCompatibleRange — a sized forward range whose value type is triqs::stat::StatCompatible. This is the type of input accepted by triqs::stat::mean, triqs::stat::mean_and_err and triqs::stat::jackknife.

Concepts
concept	triqs::stat::AccCompatible
	Concept to check if a type can be used with triqs::stat::lin_binning or triqs::stat::log_binning.
concept	triqs::stat::StatCompatible
	Concept to check if a type can be used with various Statistical analysis tools.
concept	triqs::stat::StatCompatibleRange
	Concept to check if a range can be used with various Statistical analysis tools.

Concepts