News for hackers. Stuff that matters.

A comment on Julia vs. Python: Monte Carlo Simulations of Bitcoin Options.

Re:Julia vs. Python: Monte Carlo Simulations of Bitcoin Options by spikels on Saturday March 22, 2014 @06:01PM (#7451087)
Anyone know why the two Bitcoin Call Option Valuations get such different results: Python 361.203 and Julia 374.496? Monte Carlo will always some error but that's a pretty big difference. Can't confirm which more correct right now.
[ Parent | on HN ]
Re:Julia vs. Python: Monte Carlo Simulations of Bitcoin Options by sirsar on Saturday March 22, 2014 @10:25PM (#7451844)
This bothered me as well. That's a 3.6% error, which is far from negligible. Was the main problem too few trials?
[ Parent | on HN ]
Re:Julia vs. Python: Monte Carlo Simulations of Bitcoin Options by rawrjustin on Saturday March 22, 2014 @11:35PM (#7452000)
the problem seems to be too few trials. 10x the number of paths and the calculations for both languages converge on around 361ish
[ Parent | on HN ]