Bivariate Plackett Copula Addition

[Santymax98]

The file PlackettCopula.jl is added that includes: an instance of the bivariate Plackett Copula, calculation of the CDF and PDF functions, the random sample generator and the Spearman Rho calculation. Additionally, I added some testing these functions in biv_plackett.jl.

[lrnv]

Hey,

Thanks a lot for this proposal ! I changed a few things to comply with the current standard, but globally everything is working great. I did commit to your branch so please pull down from it before doing anything more.

One thing : Your tests are not passing on the pdf and the cdf, because julia is computing values that are more precise than what you asked for, maybe you should choose is you want to round them or something so that the tests pass ?

Moreover, where are these values coming from ? the reference ? You could write that as a comment in the test file.

I am not sure about what to do with the Base.rand() and Distributions._rand!() situation, it seems like from a copula to the other there is no clear standard, I should investigate that. But this is not a problem for this PR.

Correct your tests, and then we can move on to other models :)

[Santymax98]

I manually calculated the values ​​for each of them and tried to put in a fixed number of decimal values. I took as reference each of the functions present in the texts. However, could you take some examples or exercises to strengthen the tests a little? When you set Base.rand() and Distributions.rand!() does it give any ambiguity conflict? I am going to implement the suggestion of the tests, it is the first time I use them so I don't have much practice in it. After them, if you are interested, you could continue implementing some Archimedean copulas and try to include some extreme value copulas. El El jue, 26 oct 2023 a la(s) 10:14, Oskar Laverny < ***@***.***> escribió:

…

Hey, Thanks a lot for this proposal ! I changed a few things to comply with the current standard, but globally everything is working great. *One thing* : Your tests are not passing on the pdf and the cdf, because julia is computing values that are more precise than what you asked for, maybe you should choose is you want to round them or something so that the tests pass ? Moreover, where are these values coming from ? the reference ? You could write that as a comment in the test file. I am not sure about what to do with the Base.rand() and Distributions._rand!() situation, it seems like froma copula to the other there is no clear standard, I should investigate that. Correct your tests, and then we can move on to other models :) — Reply to this email directly, view it on GitHub <#40 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A7PTWNXC6CWG7H5XOH4CMFTYBJO43AVCNFSM6AAAAAA6QMZ6GWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBRGEYDCNRXHA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[lrnv]

I fixed the tests so that they pass, now we could merge this.

If you want to add more to this PlackettCopula, you can still change things. Otherwise, tell me so and I'll merge this pull request.

[codecov]

Codecov Report

Merging #40 (9494de8) into main (810ebb9) will increase coverage by 3.68%.
The diff coverage is 94.28%.

@@            Coverage Diff             @@
##             main      #40      +/-   ##
==========================================
+ Coverage   67.70%   71.38%   +3.68%     
==========================================
  Files          21       22       +1     
  Lines         322      360      +38     
==========================================
+ Hits          218      257      +39     
+ Misses        104      103       -1     

Files	Coverage Δ
src/Copulas.jl	100.00% <ø> (ø)
src/MiscellaneousCopulas/PlackettCopula.jl	94.28% <94.28%> (ø)

... and 3 files with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[Santymax98]

I think that for now it will only be that with respect to Plackett's copulation, the rest I will do over the course of this week. To add new copulas do I always need to create a new PR?

[lrnv]

Well, This is probably the easiest thing to do yes, to create a new PR. You might add a few of them at once if you want. This PR interface allows us to discuss and collaborate easily on the code, dont you think ?

I'll merge and close this one now, thanks a lot !

Btw, Congratulation on you first merged pull request !