Re^7: Curious find while comparing grep, map, and smart match...

[space_monk]

The output of your program is meant to be a set of unique random numbers. How it is achieved is surely irrelevant (well almost). The code produced by BrowserUk delivers the same output as the previous 3 functions in a much shorter time.

I hope this is regarded as informative and educational. TBH, I didn't think of looking at the problem in the same way as BrowserUk did until he posted his idea. It often happens, especially round here, that people look at a problem completely differently from the way you do and achieve a much better result. It's happened a few times to me as well. I believe it's called an XY problem, when people come on here talking about how to do Y when they should really be talking about X, the step that got them to Y; and its the person who first spots that who often produces the most interesting answer.

If you are arguing that what he achieved was "against the rules" in some way, and you expected each function to have to sort through duplicates, then knock yourself out....:-)

[dbuckhal]

How it is achieved is surely irrelevant (well almost).

Actually, it is absolutely relevant, because that is the core of my benchmark: comparing how fast grep, map, and ~~ process data. It is the generation of the data set that is irrelevant and probably can be replaced with BrowserUK's method, because it probably does not matter if what each of those functions process is unique, or not.

note: this posting spree has really helped me improve on my posting markups!