* Specific GEDCOM export?

[Phoenix]

Hello,

I was wondering if the following would be possible with FH, and if yes, then how?

I would like to export to a GEDCOM file all the descendants of a certain individual, plus:
- The ancestors of the partners of these descendants to a certain number of generations (e.g. only the parents of the parents, or also the grandparents, ...)
- Other previous (or later) marriages of the partners

If it is not possible through FH, then is there perpahs another way to do this elsewhere?

Thanks,

David

[Gowermick]

Yes, quite easy really, an exercise I did myself only yesterday

1. Create a query to collect all your direct ancestors, select them all from the resultset ( using Ctrl-A) and using the cogwheel, add selected records to a named list.

2. Create another query to select the people from your other criterium. Add the results to the same named list (as above) making sure you don’t replace the current contents, but add them to those already there.

Then use file/export, choose gedcom export then select the named list from above, then use >> to transfer all these individuals, into the right hand pane. Set your export criteria i.e Notes, living etc then export.

Then, from the file/project. Window select new project, create from gedcom, and select your newly exported gedom to import

Job done

[Phoenix]

Hello,

Thanks for your reply.

The thing is that I want to export everyone with a certain family name, their partners and those partners parents & grandparents (as well as other marriages of those partners).

So I think that for this, I would not need to start with myself and all my direct ancestors, but rather with the furthest ancestor with that family name down all branches, including the related people to their partners.

Would that be done the same way?

[Gowermick]

The heart of the solution lies in making a named list of all those you want to export. Once you have that, the exporting is simple.

How you populate the named list is up to you, but may be easier if you can create a query/queries to give you a resultset(s) containing all those who you want to export. (As a last resort, you could always add them manually)

For example you could create a query listing a specific individual and all their descendants, or a query listing all with the same surname, or all those born in a specific place, the options are endless!

Queries are very powerful, and once you understand them, they will be able to give you exactly who you want. Visit the knowledgebase to see what is possible

[Phoenix]

It concerns thousands of individuals so manually adding them would be practically impossible. So what I'm hoping to find is a solution where I just add some criteria and then everything is done automatically.

The thing is that I want to display my GEDCOM on a website, but only the part relevant to a certain family name. So all other branches of other family names in my family tree don't need to be in there, except for those partners and their relations as I mentioned before. I hope I'm explaining myself correctly. If it's easier I can draw it out.

I'm not sure I understand how those queries work that you mention.

[dbnut]

It concerns thousands of individuals so manually adding them would be practically impossible. So what I'm hoping to find is a solution where I just add some criteria and then everything is done automatically.

The thing is that I want to display my GEDCOM on a website, but only the part relevant to a certain family name. So all other branches of other family names in my family tree don't need to be in there, except for those partners and their relations as I mentioned before. I hope I'm explaining myself correctly. If it's easier I can draw it out.

I'm not sure I understand how those queries work that you mention.

From your later explanations, I've faced this before. As far as I can tell there is no direct way with queries, or a modest collection of these to achieve what you hope for [it would be great if someone can contradict this and point the way].

My first bits of advice are:

• Buckle down and study the queries that were suggested. You're going to need those (or similar), and more before you're finished with this.

• Get used to named lists.

• Right now, make a copy (with a different name) of your project and use only this for all your learning and experiments.

Otherwise, I guarantee at some point you will ruin the original.
With that experience behind you, my approach (or anything else that may be suggested) can be explored.

In a nutshell, that would be to start with the entire "population" and in stages, "snip off the unwanted bits".
Just how you best begin that process depends on whether the branch of interest is in the majority (harder) or minority (probably much easier).

In outline, with the easier minority case:

• Look for the earliest (topmost) connection with "the masses". That should be a marriage or parentage, obviously.

• From that point outwards (maybe that spouse's or parent's spouse/parent/etc), decide where to snip the linkage. A gross but easy way to do that is deleting the linked person's record altogether.

• Run a query including "Relationship Pool" (yep, time to learn another neat trick), noting which pool your relevant family members belong to.

• Check if there is still only one pool (apart from maybe a few unconnected stragglers in your data), in which case there are other links that still connect your family to the wider set.

• If so, you need to identify those links (may take a while) and snip carefully, as you did in the first place.

• Go back to Run a query including "Relationship Pool" and repeat this whole process until you can see a "stand-alone" pool that looks about right for your target family (still containing other unwanted "remote" members).

[You'll maybe need help with relationship pool, I forget how that's done]

Only then, run another query selecting only records with the relevant pool number, then (as explained in an earlier post) that's the set of records you add to a named list. Except at this stage the aim is to create a "mini project" where you will be able to focus on the relevant family with its "baggage", and deal with that.

In this (hopefully much smaller) database, you're still going to need to trim manually, I suspect, as it seems impossible to use queries to identify all the right relationship cut-off points. That seems to be more of an ad hoc decision, anyway.

FWIW my own approach to that task is visually with diagrams. I'd need to start with all descendants of the most senior person, inspect all the branches in turn, and at every level snip the unwanted spouse or child(ren) where necessary.

BTW a huge assistance in that process would be to make the diagram boxes of people with the "right" surname stand out with coloured "fill" (see how to modify a diagram "text scheme". You should be able to detect that, zoomed out to even the full descendant tree.

Once you gone that far, it may be helpful to run another relationship pool query to identify "what's left". Put those in a new named list, export to yet another (even smaller) working project.

Whether you do that or not, it's time to focus on sideways and upwards relationships by marriage. Another manual snipping task.

Sounds grim. It is. But the rewards are great in terms of the skills you'll learn, which will certainly come to the rescue in your later work.

As a footnote: Sorry if I sound patronising when I'm only trying to be helpful. Sorry if I've disappointed with a bleak picture. Sorry if I've missed some clever techniques (forgotten or never learned). And sorry to everyone if I sound authoritative when my experience is otherwise very patchy.

[tatewise]

Some of what Paul suggests is worth noting but goes a bit too far.
It may certainly be worth working on a copy of your project while developing the process.

You will need a custom Query that uses the Rows tab and its Relations tab at the bottom.
It can be based on the Query > Relatives and Relationships > Descendants but needs modifying.
See FHUG Knowledge Base An Overview of Queries for advice.

Firstly, on the Rows tab, untick Inc. original individual and tick Inc. spouses of relatives on the right.
Then click the Update button below.
That will list all descendants and their spouses of your chosen starting person.
It is important not to include the starting person, otherwise, the following Row Conditions will add all your tree.

Now add a new Row with Condition: Add if... Relationship: Ancestor Max generations: 2
Also, select Anyone in the current result set and click the Add button.

You can continue adding Rows for Ancestors and Descendants of Anyone in the current result set and Inc. spouses of relatives until you have enough relatives.

Lastly, on the Rows tab, General tab, enter Add if... %INDI% is and tick Parameter with label Starting Person so that your starting person is include in the Result Set.
You have not said whether this starting person's spouse and their ancestors are to be included or not, so please clarify.

Refine that custom Query until you are satisfied that it lists everyone you want to be included in the website.
Now when you use the Export > GEDCOM File... command the Select... button allows you to Add using Query... and you choose your custom Query.

Depending on what website product you are using you may also need to use the Export Gedcom File plugin.
It may also be easier to work in a copy of your main project and use the File > Split Tree Helper.
So please let us know how you are building your website.
Alos do you need to be careful about excluding living person details or other private details?

[Gowermick]

I fully understand what you want, and the way to get there is via queries and a named list.(FYI the comment about adding people manually was said in jest)

This is exactly what I did to create my own websites, listed below, one contains everyone in my tree, and the other shows just my direct ancestors, with all the wives and cousins stripped away.

Queries, and how they work are explained in the knowledgebase, and written by people more experinced than me, so may I suggest you read that first, then come back if further explanation is needed.

[Phoenix]

Thank you everyone for your replies and your help!

I will briefly explain how I work: on a daily basis, I use the Geneanet website to update my family tree information, as I find it very user-friendly and it allows me to access it from any device with internet connection.

The idea is that I do an export in GEDCOM-format from Geneanet, import it in FH and there export again only that data which I need - again to GEDCOM, which is then imported into "HumoGen" to display on my own website.

Geneanet allows to export partial data from your family tree, but the possibilities there are much too limited for what I hope to achieve here. Hence I was thinking to use FH instead for this. It also means I don't need to worry about corrupting or losing my data, the original data is always preserved on Geneanet.

While I don't use FH on a daily basis, I have used FH in the past to convert an Excel to GEDCOM (a project completely unrelated to all this I'm trying to do now) - which was the main reason why I purchased the FH program. And since I have FH now anyway, I figured I might as well use it for this if possible.

My intention is to do periodical updates on my website, so the process of pruning data not relevant to this website I want to import it to, should not be something that takes me hours to do everytime. If the process is a relatively easy one to do on a regular basis (even if it means it will take me some time for the first time to figure out how the query system works) then that would be ideal.

The family name I want to display everything of is DOM, but an additional issue is that I have in the meantime also branches of DOM families which do not share the same ancestor as mine (in other words: no relations of mine as far as as is known so far) but are in there because (1) two people with the same family name but of unrelated branches happened to have married each other in the past, or (2) there is coincidentally a common ancestor but with a different family name (and 3, simply no connection at all but people I once thought to be part of my family and a correction turns out they no longer are).

So yes, this makes things even more complicated, I know.

So, pondering on this further myself, I am thinking if it wouldn't be easier to approach this a different way:

Would it be possible in FH to decide that a specific person in my family tree should be removed, and automatically remove all their ancestors in the process? And by that I really do mean remove those individuals from my family tree altogether (as I said, the original is on Geneanet so no data is really lost no matter what I do with it in FH), and then I just export all remaining indivuals to GEDCOM and problem solved?
If this is perhaps a stupid question with a simple answer, I apologize, but as I mentioned already, I don't really have experience working with FH on a daily basis and so far used it only for converting data for a different project unrelated to this.

Sorry for making it perhaps more complicated than needed, and thank you for your help - it is much appreciated!

[tatewise]

You ask: "Would it be possible in FH to decide that a specific person in my family tree should be removed, and automatically remove all their ancestors in the process?"
Yes, that is easy.
Use the File > Split Tree Helper... command.
Select Delete all Individual records found by the Individual query below.
In Individual Query: select standard Query Ancestors.
Under Other Records optionally select the Delete ... Records left with no links to them.
There are other delete options you can choose if you wish.
The Help button explains the whole process.
Finally, click the Perform Tasks button.

Using the Ancestors Query will include the chosen person, their spouse, and all their direct line ancestors.
However, it will not include descendants of those individuals who will become 'orphaned'.
So you may need to develop a custom Query, along the lines I described earlier, to include all their branches.

Note that if it is always the same branch that needs removing then instead of a Parameter Label prompt the Query can use a named person as the root of the branch, but that assumes they always have the same Record Id after import from Geneanet.

If multiple branches need removing then the Query can prompt/name multiple roots.

[jelv]

Would turning it round and creating a named list of everybody to be excluded from the extract be an option? This could be done using queries. Then when you come to do the export, first add everyone then remove those not to be exported using a query.

[tatewise]

There is no particular advantage of exclusion options over inclusion options.
An important point is that any Named List approach must be compiled after the FH import from Geneanet, because the new Project will have empty Named Lists.
If a suitable Query can be created then that is available to all Projects and can be used directly by the various record selection dialogues without needing the Named List as an intermediate step.

[Phoenix]

Hello again,

I have in the meantime tried some things. The Split Tree Helper command does indeed do a nice job to just delete an entire line of ancestors not needed on my website. But indeed, as you say the other descendants of those deleted ancestors still remain, so a query would work better.

The query I need should remove all descendants of a particular non-essential ancestor, EXCEPT the ones I want to keep and THEIR descendants.

Now I've followed your (Mike's) advice there, and did as follows (so I hope I did that correctly so far):

Go to View > Query > Relatives and Relation Ships > Descendants

Then "save as custom query" and gave it a name.

I set a certain ancestor as starting person. I don't know if, in this situation, "inc. original individual" should be checked too as that ancestor should be deleted himself too.
I checked "inc. spouses of relatives"

I click update, and then click add.

Then in the "relations" tab at the bottom, I choose "exclude if", relationship "descendant", and select "this individual" and choose the furthest ancestor that should not get deleted. Then do update again.

When I look in the "results" tab then, I see that the ancestor who should NOT get deleted is STILL listed in there, as are all his descendants. So I've obviously done something wrong, but what?

When I click on any of the two defined query filters then, I see that the field next to "this individual" is empty each time.

So I'm not sure what I'm overlooking here?

Thanks!

[tatewise]

The answer lies in the detail, so to help you we need to see exactly what Rows tab Filters you have defined.
Otherwise, we are somewhat stabbing in the dark.
Please post a screenshot of that Query tab as explained in Forum Usage Tips under Attachments and Taking Screenshots.

Also, explain in your own words what you hope each Row Filter should achieve.

[Phoenix]

Hello,

Thanks for your response and help.

I'm attaching two screenshots, one for each filter I set up.

The first "add if a descendant"... is intended to list all descendants of an ancestor, so a whole family I want to remove, except for the individuals I want to exclude in the second filter.

The second, "exclude if a descendant"... is intended to specify which individual, as well as his spouse(s) and descendants, should NOT get removed and therefore be an exception to the first filter.

As you can see on both screenshots, the "This individual" field at the bottom appears empty. Yet, I have made sure to specify the individual I want in there for each filter. When I specify the individual, the field shows it at first. When I then go to the other filter and then look back at the first filter (or vice versa), then the field shows empty every time. For both filters.

The "result set" tab does show a list of 289 individuals, who at first sight seem to be a correct list as far as the first filter goes. But the second filter does not seem to have any effect on that list, no matter how often I try to click "update".

Thanks.

[tatewise]

I have deleted one of your screenshots because they both show the same Query.
You misunderstand how a Query works.
When you Run Query using the blue circular & white triangle icon, all Rows tab Filters are executed in the sequence shown.
In this Query, that will prompt for the Starting Person and both the Filters operate on that person.

When I run an identical Query it lists no records at all, which is what I would expect.
The Add if... filter finds all descendants of chosen Starting Person and adds them into the Result Set.
The Exclude if... filter excludes exactly the same descendants from the Result Set and so will always be empty.

The Rows tab Add and Update buttons make changes to the Filters, but have no effect until you Run Query.

To change a Filter, first select it, then make changes in the bottom panel, and finally click Update.
The changes will then appear in the selected Filter in the top panel.
To see the effect on the Result Set you must use the Run Query icon button, which as I said runs all the Filters.

[Phoenix]

Yes it seems I have clearly misunderstood how the Query works then.

But I'm not sure how I can delete all descendants of a certain ancestor then, while at the same time excluding a certain individual and his descendants.

Because if I do 2 separate queries, then in the first query the person I don't want to be deleted (and his descendants) will already be removed.

Does this mean it is not possible to do what I want to do in a query?

If that is the case, then I would have to manually remove the parents of the descendant I don't want to remove, and then run the initial query again, to make sure the query would not affect the person I want to keep as I would have disconnected him from the family to be deleted.
But that would be alot of extra work, so if there is a way to do this easier using a query then that would be much preferred (since I would have to do this not for only one family, but for a few more).

Thanks!

[tatewise]

I'm getting somewhat confused about your objectives.

Last Thursday you asked: "Would it be possible in FH to decide that a specific person in my family tree should be removed, and automatically remove all their ancestors in the process?"

Now you are asking: "How can I delete all descendants of a certain ancestor?"

So which is it?
A)
Do you want to choose a person, delete their entire ancestral branch, and keep only their children & descendants?
i.e. delete all direct line ancestors (parents, g-parents, etc) and spouse/descendant/ancestral branches of those ancestors.
B)
Do you want to choose a person, delete all their entire descendant branches, and keep only their ancestors?

[Phoenix]

Sorry for any confusion.

I did indeed ask to just remove all the ancestors of a certain individual, which is easily enough done with the splitter command.

But then I realized that, for some families, I have not only listed the ancestors but also brothers and sisters of those ancestors, and their descendants in turn. By just removing ancestors with the splitter command, I would still have a bunch of other descendants from that furthest ancestor. In other words, I wouldn't remove everyone I need to remove through this method, which is why I went to look for a Query solution instead.

So what I want to do his:

Of a certain individual X, remove:
1) All his ancestors and their spouses
2) all the other descendants of those same ancestors (e.g. the brothers and sisters of those ancestors, their children and their descendants and so on).

Of that same individual X, keep:
1) The individual X himself and his spouse
2) The descendants of individual X

If my reasoning is correct, then I should have deleted the entire family of individual X (all ancestors and their descendants) except for individual X himself and his descendants, meaning that I won't have any loose branches floating about anymore of descendants that weren't deleted yet even though their ancestors were.

I hope my explanation makes sense. If not, let me know and I will try to draw it out as an example instead.

[tatewise]

OK, understood.
It does get a bit more complex than it first seems.
So your custom Query needs the following Filters to select the ancestors to be removed...

1st Filter includes all direct line ancestors but not the starting person or their spouse:
Add if an ancestor of ["Starting Person"].

2nd Filter adds all the descendants of those ancestors:
Add if a descendant of anyone in the current result set. Include relatives spouses.

Unfortunately, that adds the Starting Person and their spouses & descendants so a 3rd Filter is needed:
Exclude if a descendant of ["Starting Person"]. Include original individual. Include relatives spouses.

4th Filter adds ancestors of the descendants left from above:
Add if an ancestor of anyone in the current result set. Include relatives spouses.

Depending on the complexity of the branch you may need to repeat the 2nd to 4th Filters a number of times.
i.e. How many generations up and down that branch includes.

I suggest you construct the Query by adding one Filter at a time and running it to see which records get included.
The final Result Set lists who you want to delete using the Split Tree Helper.

That also relies on this ancestral branch not having any common ancestors with the section of the tree you wish to keep.
In that case, there is a Plugin that will help.

[ADC65]

I might be misunderstanding the requirements here as they seem to have changed a lot throughout the thread. The latest requirement seems to be just export a person, X, their spouse, and any of X's descendants.

Of that same individual X, keep:
1) The individual X himself and his spouse
2) The descendants of individual X

Surely this is a simple Descendants query, adding two filters - "Add if a descendant of" and "Add if a spouse of". Save the query, use File > Export Gedcom then Select > Add using Query > [Saved Query Above], and only the Individual, his/her Spouse and his/her descendants are exported? I don't see the need to mess around with deleting people if you can just extract who you need. Or have I missed something?

[tatewise]

Adrian, you are missing that the Descendants and Ancestors queries only include direct line descendants or ancestors and not the entire descendant or ancestral branch. As usual, some simple experiments on your own tree will show that.
i.e.
They show the same as a Descendants Diagram or an Ancestors Diagram.

For example, the Descendants cases do not show any ancestors of those descendants, such as ancestors of spouses, not spouses of spouses and their children or ancestors. i.e. All the relatives via marriage.

[Phoenix]

@ Mike:

I think I have been able to configure the query correctly with your instructions. I now have 225 results (instead of 289 before) and at first glance, it looks more like what I need.

Just a question: what do you mean by repeating the 2nd to 4th filter a number of times?
Does this mean I need to push the "run query" button multiple times? And do I need to just disable the first filter for that? Because if I try that, I have zero results.

Also, would it not be possible to just select all results and hit delete? Instead of using the Split Tree Helper I mean.

As for multiple common ancestors: that is not the case here. I only have people from a to-be-deleted family that happened to marry with someone, whose brother or sister married with the family that is to remain.


@ Adrian: it was my initial thought, but the problem is that I also will work out family trees of namesakes that do not share a (known) common ancestor with mine, and those would then be left out, which is what I'm trying to avoid.

[ADC65]

Hi Mike, I must be being a bit dense this afternoon, but I don't see the difference between "direct line descendants" as you state and "The descendants of individual X" as requested ... where would these differ? I have tried this out on my tree as suggested and can't see where there would be a difference.

@ Phoenix: Again, I must be missing a point here, sorry. You are asking for "The descendants of individual X" ... what does that have to do with whether they have a common ancestor to you?

[tatewise]

@Pheonix: I mean you must enter the 2nd, 3rd & 4th Filters again as the 5th, 6th & 7th Filters and again as the 8th, 9th & 10th Filters. So you end up with one Query with 10 Filters.

Each repetition caters for another up & down generation branch.
Remember that when adding Ancestors only direct line ancestors are added for the people in the Result Set.
When you add their Descendants that may add some more people to the Result Set.
If those extra people have linked ancestors then they must be added and the whole cycle repeats...

No, you cannot just select the Result Set and hit Delete.
You would have to select the Result Set and copy those records to a Named List and then you can bulk delete them.
See FHUG Knowledge Base Delete One or Many Records.

You must also consider any other types of records that may be linked to the Individual records being deleted.
e.g. Linked Source, Media, Note, etc, records and any records that may be linked to them.
That is where the Other Records section of the Split Tree Helper is useful.
Otherwise, you would have to find all those orphaned records by hand.

@Adrian, the primary requirement if you read the entire postings revolves around entire branches of the tree.
The primary specification is:
Of a certain individual X, remove:
1) All his ancestors and their spouses
2) all the other descendants of those same ancestors (e.g. the brothers and sisters of those ancestors, their children and their descendants and so on).
Where the "and so on" is crucial to meaning not just direct line relatives but the whole branch.
Pheonix should have said the same for the Descendant branch. However, that part of the requirement is redundant.