* How to find the oldest families

[Ron Melby]

Lets say I have a table of names:

Ames, Annabelle
Zufelt, Richard

they have one thing in common, they have a source @s1944@ which happens to be FindaGrave
now in my PAGE I have http://www.findagrave...........112123 and so on
in my note, I have section plot and all that rot
So, now I have some magic with dates
and my file is:
I1 Ames, Annabell (1870-1899) http://www.findagrave................
I99 Zufelt, Richard (1902-1946) http://www.findagrave.............
(I can build that table stealing from List all citations for a source plugin

Now we are at families

I read all the FAMS and FAMC records
and end up with something like:

bbbbbbbbb FAM 102 Husb I111 FAMC FAMS
bbbbbbbbb FAM 102 Wife I312 FAMC FAMS FAMS FAMS FAMS
bbbbbbbbb FAM 103 Child I776

lets say for this discussion I put all the marriage dates in the blank field
if there is not a marriage date I will take husb birth date and subtract 20 if no husb then wife date, if no wife date then oldest child and subtract 1 year down the line adding -1 to each visit of a child for year subtraction. I realize this will not comprehend ALL the various methods of playing park the cadillac in the garage, but should yield a date for better or worse for every fam "MARR".

So sorting it from oldest to recent, I read that table,
put out:
I99 Ames, Richard (1902-1980) http://www.FaG...
I212 Hendershott, Amelia (1912-1986) http://www.FaG... <first wife
I100 Ames, Andy (1933-1999) http://www.FaG...
I986 Bilabong, Thistle (1936-1966) http://www.FaG ... Sons wife
I999 Ames, Fredrika (1945-1945) sons daughter
I101 Ames, Milburn yadda yadda yadda blah blah blah <Richards 2nd son
and so on.

As I read thru Each Fam, I delete the read from fam gob from each iteration of offspring, wife and child
If a husband or wife is missing, or if the indi record doesnt have a FaG record then it writes a blank line after the indi number in its proper place under the indi in the output table.
when I am done with the original OLD family gob from the oldest to recent table, I delete it
read next family gob and repeat until there are no family gobs to delete.

Then done, output

can anyone help me with the logic and point out my problems here? or have a better way to do this by families?

[tatewise]

Ron, it is the same recurring problem.
You obviously know what you are talking about, but we don't.
So please summarise, in just a few lines, your objectives, and what problems you need solved.

How can we comment on your logic if we don't know your objectives?

Is it that you are having problems sorting the details into date order?

[Ron Melby]

How do I find the oldest (in time) families?

I happen to know that Lovina Churchill is my oldest record in my file, she was born 1775 and she has a family.
I want to write a file that has her and her husband and family down to her last descendent who is alive today in order, oldest to newest with the information if there is a link to FaG in my file.
when they are exhausted, the next oldest family
until done with my entire file.

[Ron Melby]

here.

https://www.findagrave.com/memorial/204 ... les-sibley
lets say he is the oldest in the file.

go to his record write his wife, his children, their wives and children one line, in order all the way down until you can find no more. assume you can click the links all the way down and remember where you left off.

so when you get done with Eldah go on to john and when you are done with that we will go to the next oldest family, the Hooks in St. Mildreds in Tenterden, Kent, down to my children and then we go to my fathers side.

[tatewise]

The 1st question is why do you want to do that?

The 2nd question is what are your problems?

The 3rd question is how do you want to handle common descendants?
e.g.
Let us say the oldest family is your paternal grandfather & grandmother.
So you will list them, and their children, including your father & mother, and their children, including you.

Then the next oldest family may be your maternal grandfather & grandmother.
So you list them, and their children, who includes your mother, but she was already listed, so is she omitted along with your father, you, and your siblings, who were already listed before?

[Ron Melby]

1st answer. Clarification please, I dont understand you. Does it affect the algorhithm?
2nd answer how do I find the oldest families? Hence the title of the thread. problem I am facing

3rd, when they are common it runs out to the end from the first common INDI all the way thru, before moving on to the next record, or.........I am open to suggestions. How is it handled in a register format?

I guess alphabetically would be alright, and so much easier, but I am trying to learn manipulation of families, so I can do some things advanced in network stuff which I have mentioned before. and to be able to interrogate, order and strip apart family records and then continue on at the next (for lack of a better description) uplevel, where I left off.....has some

[PeterR]

In any one family line, the oldest family is the one where neither husband nor wife has an ancestor (even if you don't know any of the dates). Does this help at all as a starting point?

[Ron Melby]

well, yes, and I thought of that.

then the algorithm gets beyond me.

Read FAMC Records, where there is none its a treetop,
trivial as one pass thru the file, these are 'the oldest' records I have for any family line.

I have read many FAMC and FAMS records on that pass that I have ignored.

So, a second pass to build down somehow?
so a two pass at the file? read the OLD FAM and run down their threads?

There is the difficulty, how to key or position them in the table so I just read down and deal with them as they come up, and where there is a cross relation, pick up there and read to end. then pick back up at the next CHILD or whatever RECORD, because I have several fellows who married two sisters in turn (and one that didnt marry lol) and other such shennanigans.

so If I end up with FAMC records left over and INDI FaG matches: error, or indi records left over; error.

My file is around 3500 now with probably over half having a FaG dispostion of some sort, since I come from some pretty old people. LOL.

[tatewise]

OK, after a little thought, this is how I would tackle it.

1. Search Family records to find those 'tree tops' without any Ancestors.
   i.e. Each spouse (HUSBand & WIFE) either does not exist or has no Family as Child (FAMC) parental link
   Determine a Date Point for each such Family using the first of the following:
   1. Date of Marriage Event
   2. Estimated Birth Date of a Child or Spouse see plugins:code_snippets:estimated_birth_dates|> Estimated Birth Dates (code snippet)
   3. Use an arbitrary Date such as 1 Jan 1000
   Insert the Family record and its Date Point into a table to allow for sorting
   e.g. table.insert( tblTreeTop, { Family=ptrFam:Clone(); Date=dptDate; Descendant={ }; } )
   See later for use of Descendant table
2. Sort table: table.sort( tblTreeTop, function(tblFam1,tblFam2) return tblFam1.Date:Compare(tblFam2.Date)) > 0 end )
   This relies on the Date table entries being Date Point Objects not Date Objects
3. Loop through the tblTreeTop now in date order: for intIndex, tblFam in ipairs (tblTreeTop) do
   Then for each entry lookup each tblFam.Family member Individual record using a dedicate Lua function
   i.e. Each instance of HUSBand, WIFE, and CHILd (there may be more than one of each)
   For each of those Individual records perform the following:
   1. If already added to another tblTreeTop entry then skip them: if tblUsed[intRecId] then return end
   2. Use their Record Id to show they are in a tblTreeTop family: tblUsed[intRecId] = true
   3. Insert them into the tblTreeTop using table.insert( tblFam.Descendant, ptrIndi:Clone() )
   4. For a CHILd, loop through their Family As Spouse (FAMS) Family records
   5. For each such Family record invoke the same Lua function as above recursively
      This recursion automatically traverses all descendants, adding them to tblTreeTop
      The method you use will determine whether they are added generation by generation, or descendant line by line
      The benefit of recursion is it remembers where it has got to in descending the tree by using local instances of variables
4. Traverse each tblTreeTop entry, and each Descendant individual entry therein, to produce your final list
   At this point you look up the Find a Grave entry for each Descendant individual record
5. Finally, loop through each INDIvidual record to see if anyone is not logged in tblUsed[intRecId]
   These are completely unrelated people without any links to Family records.

[Ron Melby]

I have the list of FaG entries, they are in INDI records, which I can extract into a table using list source code as I noted at the outset.

rather than lookup on FaG, I will lookup from a Table. premade, thats why we had that talk about keeping a table permenant some time ago on another thread.

The gravamen of this exercise will be traversing the FAM records and juking (a Yankshire-ism for faking and stepping around grandiously, like a basketball player making moves to go to the bucket, or a football or rugby guy fancy stepping to avoid tackle) them in several ways, and keeping track of where I am in each record (I can see situations where I might have several FAM records in process at one time... the second wife of the grandson of the old record and the wife was married several times as well as an example, sort of like one might walk thru a property box-focus window to do something specific to each record in that----well----gob.

I realize this list method will have some conundrums, like you have mentioned the cross marriages (ribbons on the diagram screen) as well as the fact I may have ancestors of a non-related wife treetops that preceed my relative treetop. But you saw how long it took me to move sources from NAME to <whole record> so this may run into a couple hundred pages yet. But your outline has clarified a great deal of the processing and thank you for that.

[tatewise]

I have made a small revision at the start of my step 1. to make it clearer, and added step 5. for unrelated persons.

There are at least two special cases that any such algorithm must cater for:

1. Family records where there are only sibling CHILdren and NO parents (HUSBand/WIFE)
   A search of INDIvidual records looking only for missing Family as Child (FAMC) will NOT find those tree tops
2. INDIvidual records completely unrelated to any family
   A search of only FAMily records will NOT find those solitary INDIvidual records

My algorithm should cater for all your 'conundrums' but will produce 'twiglets' for small ancestral branches linked to any main descendant trees.

I am not clear about the significance of the FaG cases.
Do you only want to include Individuals who do have a FaG citation, or only those that don't?

[Ron Melby]

I am looking to make that a param. A run of haves and a run of of have nots as it were.