Can you provide a small sample data set that reflects the distribution of combinations in your typical real data? Of particular interest: are combinations likely to be fairly unique (hardly anyone wears the same hat for example), or does everyone follow the fashion leaders so there are a few fairly unique combinations and a lot of very similar "me too" combinations, or maybe fashions are regimented so pretty much everyone wears the same uniform with only small variation?
In reply to Re: Help thinking about an alternate algorithm
by GrandFather
in thread Help thinking about an alternate algorithm
by Limbic~Region
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |