* Query to find family branch members

[jelv]

Lorna,

The function would be used as part of two add if row filters (one for the grandparents father and one for the mother) so you would be testing the path between each individual in the file and the selected grandfather. So the expressions would be something like

=istrue(RecordId(%INDI.~FATH>%) = textpart(relationshippath(["grandfather"],,),2))
=istrue(RecordId(%INDI.~MOTH>%) = textpart(relationshippath(["grandfather"],,),2))

[jelv]

I've given the query I created based on Mike's suggestion a good bashing.

I've found some omissions, where linking individuals were married multiple times. But out of over 2000 individuals in my file only a handful were missed.

(I think using my suggested new function would probably pick them up)

[tatewise]

Have you tried adding a few more of the Rows tab repetitive triplets?

I wonder if there needs to be more of those triplets than the number of multiple marriages.

[jelv]

Adding more rows didn't help. However...

At first sight adding

Add if a spouse of anyone in the current result set. Include relatives spouses.

before each of the exclude rows seems to have done the trick.

I'll need to do a lot more checking which I'm not likely to have time to do until tomorrow afternoon.

[quarlton]

I've just run John's 'All Branch Relatives' query, with the added
'Add if a spouse of anyone in the current result set. Include relatives spouses.'
The results are pretty good.

Based on a manual process I get a result set of 365 people.

The query returns 304, leaving 61 'missing'

I checked out the missing people and found that:
34 are descendants of the root person (my grandfather)
e.g. children, grandchildren etc. and associated spouses [Excludes his daughter, my mother and me].

The remaining 27 are no direct relationship to the query root
e.g.
No direct relationship found between Alan James CURNOW and James CARNEY.

Alan James CURNOW is father-in-law of Patricia Gertrude CARNEY and Patricia Gertrude CARNEY is daughter of James CARNEY
[i.e. his daughter's spouse's father]

Many thanks for your efforts

[jelv]

The modification I'm testing will I think pick up that last example.

[tatewise]

It is questionable whether Patricia Gertrude CARNEY daughter of James CARNEY [i.e. his daughter's spouse's father] are members of Alan James CURNOW's grandfather branch.

IMO James CARNEY is an equivalent 'grandparent' in his own right.

There are potentially a large number of grandparents each with their own branch...
You have at least 4 grandparents (possibly more if you were adopted or fostered).
Your wife has a similar 4 or more grandparents.
Your and your wife's uncles and aunts have parents in the same generation as your grandparents and IMO James CARNEY falls into this category.
The spouses of any other married descendants (children, grandchildren, etc) also have ancestors in the same generation as your grandparents.

[LornaCraig]

IMO James CARNEY is an equivalent 'grandparent' in his own right.

Sorry Mike, I interpret this differently. James CARNEY might fit in almost anywhere, but we don't need to know where. I think the point is that he himself has been correctly included in the result set, having been 'caught' by the existing row filters, but but his daughter’s father-in-law has not, because there have not been enough iterations of the repetitive triplet of Row filters.

Incidentally I see that quarlton says some of the 'missing' people are descendants of the root grandfather, so it seems that he does want the grandfather's descendants to be included in the result set. This means some of the previous suggestions won't work.

[quarlton]

James Carney (my grandfather) is the person that I have been using as the root person.

It seems (to me) a little peculiar in who is returned and who isn't.

For example the query returns Patrick Bird who is described as

No direct relationship found between Patrick BIRD and James CARNEY.

Patrick BIRD is grandfather of wife of Bernard GUINNANE and Bernard GUINNANE is cousin of James CARNEY

It would appear that the query goes across and up from the root quite happily, but as it doesn't go down from the root to his children etc. , it can't then go up on the side of their spouses.

When I first posted the question I had no idea just how complicated it was going to be

[tatewise]

Sorry, I got the relationship the wrong way round. You are correct that this Query does not explore descendants of the chosen grandfather. The analysis below explains why.

I'm still not entirely sure whether Dave wants to include descendants of the chosen grandfather in the branch.
If the answer is 'yes' then that opens up another bag of worms.
Does it include the spouses of such descendants and all their ancestral and descendant relatives?
If so it includes Dave's parents, his wife, and all their ancestral branches, which includes everyone in the pool.

The rules for older generations in ancestral branches of a grandparent require all branches to be traversed.
If the same rules are applied to descendant branches of a grandparent then the whole pool gets included!

[quarlton]

Thanks Mike.

Yes, I can see now why it might be very difficult (Impossible?) to get what I am looking for.
i.e. All connections to the root (my grandfather) except for a given child (my mother)

If the query expanded down from the root by any child, then by default it would pick up any siblings of that child which would then include my mother and thus the whole of the tree.

The only thing that comes to mind is, if the query could go down from the root to pick up the relatives of the childrens' spouses, then we could use Lorna's suggestion of temporarily unlinking the unwanted child (my mother in this case) to prevent the query working it's way back up the Simpson tree.

I'd like to thank everyone for their input.

I had initially expected someone to point out that I was being a muppet and that I was missing something obvious:-)

Dave

[LornaCraig]

The rules for older generations in ancestral branches of a grandparent require all branches to be traversed.
If the same rules are applied to descendant branches of a grandparent then the whole pool gets included!

Yes, I think the answer is to start by using the query to select the ancestral branches, traversing in all directions, and then create a separate list of descendants to be added. The starting point for this second list would be the descendants query, with any required 'relatives of in-laws' being added manually.

[tatewise]

Consider the concept of unlinking the branch that is not required as suggested by Dave.
Although his mother is unlinked, his father is still linked, so starting with his grandfather as the root would not only include all his paternal grandfather's branches, but also include all his paternal grandmother's branches!
Thus as I said, if you try to include descendant branches of a grandfather it is a whole extra bag of worms!
So you would have to prune off several branches to achieve what you want and it still relies on there being no marriages/partnership between grandparent branches.

IMO to obtain discrete branches for each grandparent, you must focus on ancestral branches only.
The descendant branches add little useful information and are essentially the same people for every grandparent.
This is very similar to what Lorna has just proposed.

[quarlton]

Mike
It is my mother's father that I was using as the root person.
So I thought that by unlinking her from her parents there would be no connection to my father and thus the Simpson family.

[LornaCraig]

It is my mother's father that I was using as the root person.
So I thought that by unlinking her from her parents there would be no connection to my father and thus the Simpson family.

But other descendants of your grandfather, from your mother’s sibling(s), are linked to your maternal grandmother. So all her relatives would be included.

IMO to obtain discrete branches for each grandparent, you must focus on ancestral branches only.

Yes, when I suggested using the unlinking technique it was to focus on ancestral branches. Unlinking the grandfather from his 'family as spouse' unlinks him from his spouse's relatives and also from his descendants, leaving him with his ancestral branches and all their linked 'add-ons' in a separate Pool.

[tatewise]

OK, but does your mother have any siblings who will still be linked to her father and thus her mother?
If so then her mother (your maternal grandmother) will have her ancestral branch included along with all the sibling & spouse branches.

Even without siblings, the marriage link between your maternal grandfather and maternal grandmother will result in the inclusion of both their branches. So that link also needs to be broken.

The way to envisage who gets included is to imagine who will be in the same Pool after breaking the link.
So after unlinking your mother, her father & mother and her siblings & their spouses will all be in the same Pool.

This is much the same as Lorna is saying but using other wording.

[quarlton]

Thanks Mike & Lorna, I can understand the issue now.
I hadn't thought about the children going back up via their mother and into her branch.

[jelv]

I've completed testing the updated version of the query (attached) - it is now picking up the more obscure connections. I've also added columns for the birth and death dates.

It selects everyone related to an individual root person EXCLUDING:

• their spouse and anyone related via the spouse
• their children and anyone related via the children.

[quarlton]

John

The new query is picking up an extra 15 people than the first query.

However, it has managed to make a connection to my father's 2nd wife, her son and parents

Helen (Ethel R) AINSWORTH is wife of Edwin SIMPSON and Edwin SIMPSON is ex-husband of daughter of James CARNEY

[LornaCraig]

John, that looks good. I can confirm that, with the exception noted below, it produces the same results as using the ‘unlinking’ method, even when there are some remote connections where the chain contains several links by marriage.

BUT, may I suggest that the description of what it does should be clarified?
You have stated that

It selects everyone related to an individual root person EXCLUDING:
• their spouse and anyone related via the spouse
• their children and anyone related via the children.

This is only true if there are no cases of intermarriage between the root person’s ancestors and their spouse’s ancestors. I have tested this and if there is any intermarriage the spouse and some of the spouse’s relatives, including the descendants of the root and spouse, will be included. (Although not, admittedly, the entire pool. That seems to have been avoided.)

[jelv]

I think it could apply to children as well.

Updated description:

It selects everyone related to an individual root person EXCLUDING:

• their spouse and anyone related via the spouse*
• their children and anyone related via the children*

*These may be included if there is intermarriage between an individual in the branch and relatives of the spouse or children

Does that look better?

[LornaCraig]

Yes, that's better.

[jelv]

I've updated the attachment to my previous post with the version including the updated description.

Could an admin delete the attachment on viewtopic.php?p=143894#p143894 and add a note to look later in the topic to find the latest version please?

[LornaCraig]

Done.

[arthurk]

Thanks for this, John - I think I'm going to use it instead of the All Relatives query that I mentioned near the start of this thread, but I didn't have the patience (and probably not the skill either) to work it out for myself.