uploaded by Private User on May 15, 2011

Comments

  • Private User May 15, 2011 at 2:47 AM

    This graph is a side effect of something completely different, but I find it fascinating.
    I have chosen a person with the highest inbreeding coefficient (calculated using Sewell's formula) and have tried to visualise amount of repetition. It works by trolling through ancestors and once a duplicate is found, marks all the ancestors above that one in the same colour. What happens, though, is that some of those ancestors appear somewhere else on this tree and this explains inserts of red and brown colour on yellow. I have checked and rechecked this and it is definitely true.

  • Private User May 15, 2011 at 2:48 AM

    If you are keen on calculating it yourself here it is
    http://www.highflyer.supanet.com/coefficient.htm

  • Private User May 15, 2011 at 2:53 AM

    Also, it also bugged me why some repeated ancestors start as a single person, while others start with two.
    Simple explanation is that in some cases siblings are ancestors, so the first common ancestor will be their parents (ie two of them)
    Probably I got everyone lost by now.

  • Private User May 15, 2011 at 5:17 AM

    Ovaj graf je slucajna posljedica neceg drugog pokusaja. Izabrala sam osobu sa najvisim postotkom srodnosti predaka koristeci Sewellovu formulu i pokušala predočiti one pretke koji se ponavljaju.
    Graf je napravljen tako da program prolazi kroz sve pretke i kada nadje jednog koji se ponavlja sve pretke iznad njega označi istom bojom. Međutim, pojavili su se neki koji se pojavljuju na jos kojem mjestu, i ako pogledate crveni i smedji umetak u žutom polju, to je to.
    Uz to nije mi bilo jasno kako to da negdje duplicirani predak je samo jedna osoba a negdje dvije, nakon provjeravanja vidjela sam da je to u slucaju kada su pretci npr brat i sestra (ali se pojavljuju na drugom mjestu na stablu)
    AKo vam nije jasno, tješite se, trebalo je i meni cijelo popodne da shvatim šta se ovdje događa.

  • Private User April 19, 2012 at 4:31 AM

    I have this program developed further, this particular graph looks a lot more interesting, will post it out soon

No tagged profiles.
Projects

Share