questions regarding the datagenrator

[josephineHonore]

Hi
I am using this repo for a project, and I have a question regarding the dataGenerator. In the class runSyntheticExp.py - runSyntheticExp.py a comment states:

"#using 0.98 as extreme case considering the limitation of random generator"

What is the limit of the DataGenertor since the fairness probability cannot be set to 1?

[KeYang0923]

Here the limitation means that the random process is not really random (always bound by the seed). It rarely generate a fair ranking like "MFMFMFMF". M shorts for male and F shorts for female.

…

On Tue, May 28, 2019 at 2:19 PM josephineHonore ***@***.***> wrote: Hi I am using this repo for a project, and I have a question regarding the dataGenerator. In the class runSyntheticExp.py - runSyntheticExp.py a comment states: "#using 0.98 as extreme case considering the limitation of random generator" What is the limit of the DataGenertor since the fairness probability cannot be set to 1? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_DataResponsibly_FairRank_issues_1-3Femail-5Fsource-3Dnotifications-26email-5Ftoken-3DAEXQOSVGLYARFD2H7O3DNCDPXVZS5A5CNFSM4HQF4Y7KYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GWJOYMQ&d=DwMCaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=ZFutW3OZwTfiycyzq_x4Jg&m=yozN1ci1ea6wA1R9OIpYHQd3GAfcz5mMrCj7fZteQm4&s=pAGai-K2ilxZkqG3IN1bpls1p2iRnFDhtFiCS6pfmiU&e=>, or mute the thread <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AEXQOSUTJIX3CEYVCZ4BCELPXVZS5ANCNFSM4HQF4Y7A&d=DwMCaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=ZFutW3OZwTfiycyzq_x4Jg&m=yozN1ci1ea6wA1R9OIpYHQd3GAfcz5mMrCj7fZteQm4&s=oWNqYvPVtSY8jO_mbiwXC-FGe_U8Uli-zk5VBtkvIW4&e=> .

[josephineHonore]

Thank you very much for your reply,

I understand that the seed determines the sequence of the output. But could you elaborate on why you chose the value of 0.98? What would happen if you set it to 1?

[KeYang0923]

Theoretically, this value should be 1. I choose 0.98 because this is the observed maximum value generated by the random process. If setting to 1, this value, using as the normalizer of fairness measure, will cause the value of the fairness measure is a bit greater than its actual value. Other than that, nothing will happen. Best, Ke

…

On Wed, May 29, 2019 at 11:52 AM josephineHonore ***@***.***> wrote: Thank you very much for your reply, I understand that the seed determines the sequence of the output. But could you elaborate on why you chose the value of 0.98? What would happen if you set it to 1? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_DataResponsibly_FairRank_issues_1-3Femail-5Fsource-3Dnotifications-26email-5Ftoken-3DAEXQOSW4LGYFQFR72CDZZRDPX2RCNA5CNFSM4HQF4Y7KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODWPY7II-23issuecomment-2D496996257&d=DwMCaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=ZFutW3OZwTfiycyzq_x4Jg&m=sgQ-Ib0enZV8vfIxi2NnYyw-doBP9xazfVvyyLVglVM&s=BsZRsmPJB0hOdYXVmxSSkz09NwxUZPnrYBCX6v8J6eE&e=>, or mute the thread <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AEXQOSUSZYVOSAW6YGZ6SF3PX2RCNANCNFSM4HQF4Y7A&d=DwMCaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=ZFutW3OZwTfiycyzq_x4Jg&m=sgQ-Ib0enZV8vfIxi2NnYyw-doBP9xazfVvyyLVglVM&s=h3kAuGir35GCiMVBliMtz2aBWDCWlFNSKC8zMBnB-jc&e=> .