
14 Dec
2014
14 Dec
'14
6:15 a.m.
Hello all, I recently wrote a Haskell program, which accesses Lists extensivly. That was okay to verify the overall idea, but performance degraded rapidly when I made the lists bigger. There was a quadratic effect and this was not surprizing, given the nature of the code. I am somewhat aware of Haskell types which offer faster access. But they all seem to assume that I know the way data is accessed when I write the type. A new access path may force me to rededign the type. What I am looking for is something which behaves like indexes in a RDBMS, i.e. to define a list-like type on an innocent data type (like a tuple) and then add indexes as needed. Is there such a thing?