Hey Steven, I think there is a way to map x <- [1; + inf] to y <- [0;1] by putting y = 1/x
But i think there is another issue to consider : for example if we want to sort : random.choice(['a', 'b', 'c'], random=lambda: random.paretovariate(1.75)), we should have a mapping between the elements of the list (here :a,b,c) to [0, 1] and the reverser mapping, having this mapping underline that we have a notion of order and distance, "measurement" between the elements, and it is not trivial to always have a notion of "order" or "distance" between the elements we manipulate, but it could be fun to have this possibility. What do you think ? Thank you ! Best regards, -- Hamza Le dim. 18 août 2019 à 11:56, Steven D'Aprano <st...@pearwood.info> a écrit : > On Sat, Aug 17, 2019 at 09:02:54AM +0200, Patryk Gałczyński wrote: > > Hi! > > > > Recently I encountered a situation when I needed to pick a random > > population sample but with specific distribution function - > paretovariate in > > this particular case. > > Let me see if I understand... you have a paretovariate distribution, and > you want to generate a sample of values from that distribution? Or > another way of putting it, you want to generate random numbers from that > distribution? > > Use the random.paretovariate() function (which I see you know about). > There's no need for random.choice or other methods to be involved. > > That might not be what you *want*, but that is what you seem to be > describing. > > > > After poking around in the random module I found out > > that none of the so-called "sequence methods" methods support custom > random > > function except for random.shuffle > > > > I would like to propose a new optional keyword argument for all these > sequence > > methods (choice, sample, choices) that will - similarly to how shuffle > does > > - take a zero-argument random distribution function of the users choice. > > Because it should be a zero-argument function, it would most likely > require > > to pass "configured" distribution function with lambda. I would see it > > being used like this: > > > > ```python > > import random > > random.choice([1, 2, 3], random=lambda: random.paretovariate(1.75)) > > ``` > The docs for shuffle say that: > > The optional argument random is a 0-argument function returning > a random float in [0.0, 1.0); by default, this is the function > random(). > > but random.paretovariate returns a float between [1.0, ∞). By > definition, every value it produces is out of range, and there's no way > of compressing it to [0, 1) without changing the shape. > > I think you need to consider more strongly what you are actually trying > to solve, go back to trying to understand the requirements before > writing the program. Because I don't think that "Choose an item from > this finite list according to an infinite distribution" is meaningful. > > > -- > Steven > _______________________________________________ > Python-ideas mailing list -- python-ideas@python.org > To unsubscribe send an email to python-ideas-le...@python.org > https://mail.python.org/mailman3/lists/python-ideas.python.org/ > Message archived at > https://mail.python.org/archives/list/python-ideas@python.org/message/5SRPE5QGTXJT3VLQFQFF4QXMVWH7DS3G/ > Code of Conduct: http://python.org/psf/codeofconduct/ >
_______________________________________________ Python-ideas mailing list -- python-ideas@python.org To unsubscribe send an email to python-ideas-le...@python.org https://mail.python.org/mailman3/lists/python-ideas.python.org/ Message archived at https://mail.python.org/archives/list/python-ideas@python.org/message/MFLJWVSOERUWQWN5JNN46N2MQW66YITO/ Code of Conduct: http://python.org/psf/codeofconduct/
