What is the shape of Wikitree ?

+3 votes
210 views

[The purpose of this message is to clarify my understanding of the matter. I hope that it is not too wrong and that it can help others in their understanding of it.]

The name - WikiTree - is sligtly misleading. When individuals trees are merged, the result is not a tree, but a "forest."

This forest has the shape of networks of profiles, linked by triangles Father - Mother - Child. Some triangles are uncomplete and part of them will remain so.

One of the reasons we love family trees is that they can be represented on a 2D plan with simple connections*. This is due to the fact that they use:

  • either one child of every family, as an ancestor of the person at the bottom of the tree,
  • or one parent of every family, as a descendant of the person at the top of the tree.

But many families have more than one child, many of them having their own family. Representing all of them on a 2D plan becomes tedious even with a limited set of generations and impossible beyond that.

I am not sure of the shape - in mathematical terms - of our networks, but my guess is that they are quite complex and that we cannot grasp them with our common sense.

--------------------------

Our ideal is to achieve a single network that would gather all profiles, but we will only get close to it, as a fraction of the profiles will remain unconnected.

A concern is whether there is already a single network connecting a vaste majority of the profiles or not. My guess is that such a network does exist, but its extend is hard to know**.

The alternative implies several networks, corresponding to two main situations:

  • The first one derives from human groups with limited connections, due to geographical, linguistics, social and/or religious differences. If these limited connections are not present here, these networks are likely to remain separated.
  • The second one implies distinct networks with duplicates in common. These networks will certainly be merged, as many already are on a daily basis.

Moving through these networks is quite a fascinating experience, allowing us to have a glimpse of how people lived their lives at differents times and in different places. Hopefully, they will develop beyond the massive Europe - North America axis and extend our understanding of how human beings live their lives in all parts of our planet.

 

* Consanguine weddings makes it slightly more complex, but there are simple ways to cope with them.

** It would take a bot dedicated to that task to know more about this issue.

in WikiTree Tech by Living Pictet G2G6 Mach 3 (33.0k points)
I've come to realize that  within my portion of the tree there are duplicate entries for members of the family. if cousins (or other relations) marry, they are shown twice in the tree.

This is not a problem, they don't need merged away, they are only  one profile each, just showing up  in two places as they appear in differing family groups.

3 Answers

+1 vote
 
Best answer

Hi Jacques,

Interesting question! I'm not an expert in graph theory so my terminology may not be correct, but here a few observations.

It is a directed graph, information flows in one direction. By information I mean DNA information here.

It is not cyclical. Despite what the song sais, you cannot be your own grandpa!

So, thinking about it in general rather than the data currently on WikiTree, it should be a single rooted tree I believe. In reality, like you pointed out, it probably is a forest due to missing information, or nodes.

Better exploring this graph of data is one of my goals. Here's an example of my pedigree (well, what it was over a year ago) without copying ancestors due to "consanguinity". So each ancestor only shows up once, no matter how many branches the node may belong to.  https://wikitree-tools.googlecode.com/hg/example_tree.png

Even sticking with just the pedigree and not taking siblings, cousings and such into account leads to a messy looking graph, with edges crossing all over the place. Thinking of what it would look like to extend such a graph to all the nodes on WikiTree is a bit mind boggling...

If you have ideas on tools to help better visualize this graph we are all part of, let me know!

by Roland Arsenault G2G6 Mach 5 (58.8k points)
selected by Living Hammond
how'd you do that?

You mean that graph? I wrote some scripts in python that works on a GEDCOM download. 

See the  http://www.wikitree.com/wiki/Project:WikiTree_Apps page for more details.

Roland, we should talk more about this. I wrote a script to browse the tree and figure out how many profiles were connected in one graph, starting with a specified profile, but it was way too slow. I want to switch to using the API, but am not really good with that sort of thing (the getting started with a new technology part).
Lianne, how busy are you going to be at RootsTech? Such a script is something we could play around with a bit.
Roland, is your python script the one used in Gramps?
Roland: Hard to say. There are going to be quite a few of us this year, though I'm not sure how much time everyone will be spending at the booth. But I'm sure we could work something out! (Sidenote: So excited for RootsTech!!!)
Rob: Nope, it's all original code I wrote to teach myself how GEDCOM files are organized.

Lianne: I'm getting excited about attending my first genealogy conference as well! I'm sure we'll find some time to chat about all sorts geeky geneaology stuff...
+1 vote
Hi Jaques,

I tend to think of WikiTree starting as being a single tree (which of course grows), but I  think of the "new family trees being added" as  grafts being added on to the original tree which then grows and develops as does the tree.
by Billy Wallace G2G6 Pilot (230k points)
That's how I see it too, but I accept Jacques' point that all components are not going to be connected - ever. That really would be too much to expect. Some groups extending back to the 17thC will never be connected because the earlier information needed simply doesn't exist.
It's still better than a site like ancestry with trees never completely connected or information corrected. I am impressed with the level of care taken here and the cooperation of those who look beyond their own blood relations. I may not see the Forest for the trees, but that's because I am looking at the individual humanity, not the scheme, layout or final picture.
It's not necessarily true that the whole tree can never be connected in a single graph. Such connections don't need to be by blood. Sure, we can't get back far enough to see how the indigenous people of North America are related to Europeans by blood, but when you factor in marriages over the past several centuries, a lot of separate groups of people are suddenly connected.
0 votes
Thanks for your answers and comments.

I am not surprised that forgetting about the tree is not an easy step.

Roland's remarks are quite accurate.

I am afraid there is no easy way to represent such networks easily.
by Living Pictet G2G6 Mach 3 (33.0k points)

Related questions

+8 votes
5 answers
328 views asked Aug 7, 2023 in The Tree House by Steven Harris G2G6 Pilot (745k points)
+8 votes
0 answers
67 views asked May 2, 2021 in The Tree House by Paul Chiddicks G2G6 Mach 1 (11.3k points)
+4 votes
1 answer
+12 votes
3 answers

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright

...