Potentially ambiguous rand behaviour #33309
Labels
docs
This change adds or pertains to documentation
randomness
Random number generation and the Random stdlib
Comments
|
It was interpreted as dimensions in the past, and was changed with the new meaning (partily for performance reasons) in Julia 1.1 I think, see #25278. |
|
Fair enough. Possibly this should be clarified in the documentation to make it obvious, as it was a somewhat annoying bug to find. All I can see currently is "Support for S as a tuple requires at least Julia 1.1." |
rfourquet
added a commit
that referenced
this issue
Nov 1, 2019
KristofferC
pushed a commit
that referenced
this issue
Apr 11, 2020
* docs: disambiguate what rand((2, 3)) does (fix #33309) * Update stdlib/Random/src/Random.jl Co-Authored-By: Matt Bauman <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I ran into unexpected behaviour today calling the rand function. I expected
rand((2,))to interpret the tuple as dimensions, whereas it actually interpreted it as the set of values from which to be drawing from randomly.In my opinion, a tuple does not (and should not) represent a 'set of values', but rather be a compound data type where each element has an explicit meaning (such as with dims, where each element of the tuple represents the length of specific dimension), and hence, should not be a valid type for S.
Should this be changed?
The text was updated successfully, but these errors were encountered: