I’m confused–I thought count tree was suppose to tell you how many people are in you main tree ( blood relations/ spouse of/ descendants/ancestors etc and then how many unconnected/ not related trees you have..
can anybody help me out?
1 Like
That’s what I thought too. I understood that it was to find individuals and/or sections of the tree that were not connected to the main tree, often caused by deleting a linked individual. The disconnected people or section showing as a separate tree.
What have you encountered that calls your description into question?
I second Tom’s question. I have never seen a problem with RM’s Count Trees tool. For me, it has always worked exactly as described. Well, users have complained that the person it chooses to represent each tree is not the person they would have chosen. But I’m not sure there actually is a perfect way to choose the representative person from each tree.
My many unlinked trees come from transcribing every name in a number of early New England genealogies. I get additional names from my one name surname studies. Also frequently I happen across entries in the US census schedules when I am researching another family, but want to keep the new names I have of interest. I may not interested in those at that moment in time in researching the other family, but feel that later the family may relate to other research.
The unlinked trees do no harm. Occasionally I browse through them and actually find I can place a group of unlinked names in my main tree. At the very least they may help others who find them in the online databases where I share my research. I generally keep everything I research feeling it can help someone out later in time.
@TomH and @thejerrybryan
I looked at Mom’s file which has the bulk of people listed under her–the next largest is my godfather’s line and then a lot of other littler lines as I am more like @mscheffler collecting others in the area that MIGHT connect and saving it as it MIGHT not connect to me but help someone else…
So I get the general idea how this should work BUT
I have a Green file with 12,461 people --so NOT huge–surname statistics report shows 1331 Greens and next largest is Williams at 708 ( note surname changed for privacy)–so maybe it should be Green/Williams file..
So you understand on the Greens, this is a Mine, Ours and Theirs file–meaning that I have
Mine-the 4th g grandfather James Green to 7th grandfather Unknown Green and their siblings BUT basically very few descendants..
Ours–the cousin line that goes back to the father of Henry Green
Theirs-- other Green lines in same areas
The Williams line–basically a separate line-- no one connected to the Henry Green line—one of my Green cousins is a descendant of this line BUT don’t have this cousin connected to any of the Green lines…
So when I go to count trees-- I get a count of 7369 under James Green
( who yes has the lowest #)–so I was confused then remember that there was ONLY 1 marriage (in 300 years ) between Mine and OUR Greens (married after 1900)— so unlinked her from parents and now I get 6540 on the Williams line and 829 on James Green --which is abt right for the Williams line BUT how can James Green have 829 when
- there are only abt 8 Greens in the James Green line with very few spouses or kids–a narrative report for James’ g-grandfathers shows 20 descendants…
- the vast majority of Greens in this database are in the Henry Green line as well as all their other surnames–a narrative report for Henry’s dad shows 998 descendants
So that is why I am confused -I would have thought that after the Williams, there would have been a count for the Henry Green related line and a much smaller count for the James Green related line
Sorry so long but go back to work tomorrow after being off and may be slow in getting back–so wanted you to have as much info as possible
I don’t think I’m able to answer without being able to see your whole database. But linkages between trees can be very subtle. I have discovered this with my color coding. It’s a color coding that can only be accomplished with SQLite, by the way, not one that can be accomplished from the RM user interface because RM’s Advanced Search doesn’t support searching be relationships. But for example, two trees can be linked into one by a spouse of a spouse of a spouse sort of thing. The two trees don’t have to have a direct spouse relationship between the two trees to be linked into one tree.
For example, suppose i have a main tree and a secondary tree. If my third cousin’s wife’s first husband suddenly marries into the secondary tree (or if I suddenly notice the relationship and enter it into my database), then the two trees immediately become one.
Just speculating but you might have uncovered an anomaly in RM between or among different algorithms used for counting and reporting relatives in different parts of the program. Try this:
- Add up the populations of all the separate trees listed by Count Trees.
- Does the total match the number of people reported by Database Properties?
- Repeat after unlinking or relinking that has caused you concern.
- For a tree that you are puzzled by its count, generate a Kinship List report forr one of its members. Given the number of pages shown in the preview and the number of kin per page or an export to a XLSX file, get the total kin reported. Does it match the population reported by Count Trees?
Also, note that a Descendant Narrative does not include the ancestors and siblings of the spouses of the direct line descendants - maybe Count Trees does.
@thejerrybryan Thank you-- I understand what you are saying and it makes sense BUT I have looked and looked at common surnames between the different lines of James Green, the Henry Green line, the odds and end Green lines and the Williams lines that when all connected my count tree would have that James Green’s count would be 7136 out of 12, 461 people.
note overall count changed from previous message as I ran all the tools.
When I unlink the one person in the couple that connect the James Green line then the count for James Green drops to 788 and now the Williams line has 6348 --unlink my cousin from the Williams line and the Williams line drops to 1554 but the James Green line stays at 788–main focus is another line.
So I can see what you are saying BUT to say that there is in this database something that connects all 3 lines so that James Green would have a count tree of 7136-- I can’t see it nor find it- but will consider it and option and maybe after reading the next message to Tom you will understand my confusion..
Thanks @TomH --I think you trying to kill an old lady who just worked all night 
To answer your questions
I added up the populations of all the separate tree listed by Count trees BOTH ways and yes they do match the total of 12,461 either way.
As I told Jerry-- when I don’t change anything it says the count for
James Green is 7136 out of 12,461
when I make one or 2 changes the count for James Green drops to 788 out of 12,461 people.
When I run a Kinship list for James Green, the total kin reported is 44 people-- yes I said 44…
And when I run a Kinship List for the Henry Green line total kin reported is 1407 people..
When I run a Kinship list for the Williams total kin reported is 761 people..
So I just can’t see how James Green can have a count tree of 7136 when all are connected and 788 when not all connected
Not to muddy the water, but are kinship lists and Count Trees reports being conflated when they shouldn’t be? The concept of the Count Trees function and of a kinship list are very different. A former spouse of my cousin’s spouse is in my tree as reported by Count Trees, but is not in my kinship list. Both results are correct.
True. The tree is the network of connected people, whether by blood or by other types of child-parent links and by spousal links. A kinship report could list them all but the convention is to limit it to child-parent links and spouses thereof. I don’t recall if RM limits child-parent links to type ‘blood’ or if it includes other types.
I’m surely not explaining what I’m trying to say very well. But suppose I had a somewhat simple database where everybody was in the same tree. Then Count Trees would show the the same number of people in just one tree as were in the whole database. Suppose that in this simple database I had a cousin who had a spouse who had a former spouse. My cousin’s spouse’s former spouse would be in the same tree and Count Trees would still just be showing one tree. But if I ran a kinship report on the former spouse, there would only be two people in the report. That’s why I’m worried about conflating the Kinship list and the Count Trees report, even though I can see why the Kinship list could be a good tool to help find the boundary conditions where things break down.
I think I would approach this problem as follows. First of all, I would make a test copy of my database and work only in the test copy, because I’m going to be breaking things. I would then color code each tree a different color. It’s pretty easy. You can just color code based on the one person that is listed for each tree in Count Trees, using the Everyone in Highlighted Person’s Tree criteria. At this point, for any couple you look at, each spouse will be the same color. it doesn’t matter how many spouses a person has or how many spouse of spouse of relationship there are, Being spouses means you are in the same tree and we color coded based on being in the same tree.
The next steps do get fuzzy. We need to unlink a couple and color code again to see who changes color. Then after the unlink and after redoing the color coding, look for people whose current color doesn’t match the previous color and who have more than one spouse. The trick is to use a different color set for the second coloring. Then look for things like people who are red in color set #3 and who are not red in color set #4, where color set #3 is before unlinking and color set #4 is after unlinking. I’m certain that I haven’t thought this all the way through, but something like that should eventually solve the mystery.
That sounds logical. In my fuzzy thinking after whaling away at income tax time and re-reading what I wrote, yes, I did conflate ‘tree’ and ‘kinship list’. Sorry, Nancy, for putting you through some unnecessary hoops. Jerry’s before-after color coding comparison should be helpful.
Edit: I wonder if it’s possible that the real network could be more convoluted than the ‘Count Trees’ function indicates. Could there be fewer true ‘trees’ than it reports? I imagine that it starts with the smallest record number and tries to identify everyone in the database with some linkage to a person already linked but misses some. Then it finds the smallest record number that has not been included in a tree and starts a fresh count. That record might actually belong to the ‘real’ network but was missed and another branch of the true tree was reported as a separate tree.
Thanks @TomH and @thejerrybryan --Jerry’s color coding worked to help me understand how one ( or more) groups of people related to each other by blood and marriage can be counted in a tree that they have no direct relationship to–I tried it on Mom’s tree 1st-- where the Winks were in our count–the Winks were raised by g-aunt Kate and her hubby Herman (the Winks and Herman only had the same step mom who last name was Shoe)–well Kate’s brother s married a Huff- a Huff sibling married a Flat and a Flat married a Shoe–so we have 228 people added to the line–well Mom always did say the Winks were related and then there is the other old adage that there is only 6 degrees of separation between any 2 people..
Count trees includes all of your ancestors, their siblings and descendants plus all collateral line–so basically it works fairly well with your parents, their ancestors and descendants but NOT much help when you have several distinct lines in one file..
and yes Tom based on what I saw, it can miss some people who should have been counted in the main group BUT usually because the info we entered is not quite right-- just wish there were more trees based more on direct lines–BUT it is what it is..
Thanks for all the help.