add caching to df generating algorithms

add chaching, maybe even a generic caching function to utils? cache folder could be defined here