argh. fix one-d mass matrix momentum draws

set (dummy) mass matrix on initialization

typo for oob check

add docstrings, change how acceptance is tracked

add mass matrix implementation

add variable length option

pep8

add an autocorrelation time estimate with adaptable window

rewritten gr convergence criterion

add verbose option

update test and demo

fix docs and bug in intial lnP and grad in fortran version

demo figure

cleanup epsilon adjustment

working fortran version