[RFC]: Refactor random number generation in JS benchmarks for stats/base/dists (tracking issue)

[anandkaranubc]

This RFC proposes improving random number generation in JS benchmarks for remaining packages in stats/base/dists.

[ahmad-kashkoush]

Hi @anandkaranubc , I'm interested in closing some of these issues, are these issues aligned with project direction?
I want to assure that they will get merged right away

[anandkaranubc]

are these issues aligned with project direction? I want to assure that they will get merged right away

Hii @ahmad-kashkoush, feel free to go ahead with creating PRs targeting these issues. Probably start with one, get it reviewed, and then proceed with creating more.

[saurabhraghuvanshii]

if any one working on any issue please comment in that issue , it will be help full and there will be no 2 Pr for same issue .

[anandkaranubc]

For everyone working on the refactoring issues above, can you all please update your PRs to also include the same changes that were made in #5296?

Specifically, the set::* benchmarks in ctor need their isnan checks updated.

Thanks @G4URAV001 for pointing out this bug.

[saurabhraghuvanshii]

For everyone working on the refactoring issues above, can you all please update your PRs to also include the same changes that were made in #5296?

Specifically, the set::* benchmarks in ctor need their isnan checks updated.

Thanks @G4URAV001 for pointing out this bug.

I didn't understand what was the and bug how this solve .

[anandkaranubc]

I didn't understand what was the and bug how this solve .

Because in cases like these (where the random number generation occurs outside the benchmarking loops), specifically in the ctor set::* benchmarks, y is being initialized as an array instead of a value. That is why the isnan(y) check appears to be a bug here.

[saurabhraghuvanshii]

that is , thanks @anandkaranubc