PDA

View Full Version : Can I use Automator with Reunion?


Tim Powys-Lybbe
14 November 2006, 03:34 AM
I have a difficult query to make on Reunion: I have a marked list of about 200 unrelated people and I want to find all their descendants who are ancestors. Obviously I can go through these 200 people one by one and mark off their ancestral descendants; but this is a complicated procedure requiring the use of flags to preserve the original list. And knowing myself I will make a mistake halfway through and will have to start again.

What would be nice would be to be able to use the Automator facility that comes with Mac 10.4 (or it may be earlier, I simply don't know). Is this possible?

Gregg
14 November 2006, 07:46 AM
I have a marked list of about 200 unrelated people and I want to find all their descendants who are ancestors.If I understand what you are asking, you can do this with Find Anything.

Select the People tab and use these settings.

# of Marked Parents : More Than : 0

AND

# of Children : More Than : 0

Anyone who has one of your marked people for a parent and also has children will appear in the result list. I think that's what you wanted, but if it's not this search can be adjusted.

...this is a complicated procedure requiring the use of flags to preserve the original list.To save the set of current marked people, without using flags, choose File -> Mark Sets -> Save Marked Set...

Tim Powys-Lybbe
14 November 2006, 07:21 PM
If I understand what you are asking, you can do this with Find Anything.

Select the People tab and use these settings.

# of Marked Parents : More Than : 0

AND

# of Children : More Than : 0

Anyone who has one of your marked people for a parent and also has children will appear in the result list. I think that's what you wanted, but if it's not this search can be adjusted.

To save the set of current marked people, without using flags, choose File -> Mark Sets -> Save Marked Set...

Thanls for that and I had already tried a variation of this. But this only gets the next generation and it gets the "children of the marked people who have children". But these are not necessarily ancestors.

So I have a flag for ancestors. But adding this to the search still only gets the one generation and not the many generations of all the people between these 200 odd and the source person amd who are both ancestors of the source and descendants of the 200 odd.

The only way I can think of doing this is by an iterative process where you store the results of each search in a flag and which involves several careful actions at each iteration. That is why I asked if the Macintosh Automator tool could be used: can it?

Gregg
15 November 2006, 09:05 AM
I had already tried a variation of this. But this only gets the next generation and it gets the "children of the marked people who have children". But these are not necessarily ancestors.Sorry, I truly don't understand what you are asking.

That is why I asked if the Macintosh Automator tool could be used: can it?No.

Tim Powys-Lybbe
15 November 2006, 06:01 PM
Originally Posted by Tim Powys-Lybbe

I had already tried a variation of this. But this only gets the next generation and it gets the "children of the marked people who have children". But these are not necessarily ancestors.

Sorry, I truly don't understand what you are asking.

Let's start with a thanks for saying that Automator cannot be used with Reunion; that eliminates one answer to my problem.

The original issue was I had a list or around 200 people and wished to make from this a list of all their descendants (if any) who were ancestors of the Source person on the one database. To put some clothes on this. the 200 odd people are ancestors who are in Domesday Book - if you have ever heard of that document which was written in 1086 and which is still surviving and readable (just). The database is 80 per cent medieval people and tracks relationships from those times through to the present day. I wished to know how many people were descendants of these Domesday entries and also ancestors of the Source person. What of this do you not understand?

You suggested the following procedure:

"Select the People tab and use these settings.
"# of Marked Parents : More Than : 0
"AND
"# of Children : More Than : 0

"Anyone who has one of your marked people for a parent and also has children will appear in the result list. I think that's what you wanted, but if it's not this search can be adjusted."

Using the above in Find Anything produces a set of people who have one of the Domesday People as parents and who also have children. These people who have a Domesday parent can include a range of brothers and sisters. Some of these brothers and sisters are not ancestors of the Source person. This is why I wrote "But these are not necessarily ancestors."

Even if the above Find Anything was restricted to only picking up people who were ancestors (and had a Domesday parent and had children), this only contains the right descendants in the one next generation. But there are a maximum of 30 generations from these Domesday People to the Source person. I do not think there is a single query in Find Anything that will pick up the right people in every one of these generations. Can you think of any way this can be done in one query?

Mary Arthur
15 November 2006, 07:16 PM
I had already tried a variation of this. But this only gets the next generation and it gets the "children of the marked people who have children". But these are not necessarily ancestors.
I think I understand what you want, first mark the 200 people, unmarking all the others in the data base, and then go to the card of the person you want to know about and then calculate her ancestors (List --> Relatives) and then see which of the people in the list is marked. You will still have to follow each 'trail' separately, but this might narrow it down?

MabryBenson
15 November 2006, 07:16 PM
There is a bit of a contradiction there - to find unrelated persons who are also ancestors (if a desc of them is an ancestor, then they are too). If they ARE ancestors, then they are not unrelated.

But my thought for Find Anything is -

1. From your source person, do Direct Lines Bold (first unbolding everyone) for ancestors of the source to thus identify the source's ancestors.

2. Somehow Mark only everyone on this list of unrelated persons.

Then with Find Anything, #Bold children >0 (ie an ancestor) AND Person = Marked.

martha
16 November 2006, 12:25 AM
Thanls for that and I had already tried a variation of this. But this only gets the next generation and it gets the "children of the marked people who have children". But these are not necessarily ancestors.

So I have a flag for ancestors. But adding this to the search still only gets the one generation and not the many generations of all the people between these 200 odd and the source person amd who are both ancestors of the source and descendants of the 200 odd.

The only way I can think of doing this is by an iterative process where you store the results of each search in a flag and which involves several careful actions at each iteration. That is why I asked if the Macintosh Automator tool could be used: can it?

Tim, I am not understanding what you want to do. If family members in your database are connected to one another, then all ancestors and connected people should be able to be found. What are you searching for, exactly?

Martha

Frank Mitchell
16 November 2006, 12:32 AM
The original issue was I had a list or around 200 people and wished to make from this a list of all their descendants (if any) who were ancestors of the Source person on the one database. To put some clothes on this. the 200 odd people are ancestors who are in Domesday Book - if you have ever heard of that document which was written in 1086 and which is still surviving and readable (just). The database is 80 per cent medieval people and tracks relationships from those times through to the present day. I wished to know how many people were descendants of these Domesday entries and also ancestors of the Source person.

You may be able to do this in four steps...

1. Mark all the Domesday people -

Assuming the source no xx that you have assigned to the Domesday Book source appears somewhere on all their family cards...

Go to List/Sources (command-S)
Select the Domesday Book source no xx
In the Source Tools pop-up select Show Links to Source no xx (command-L)
Use the Mark pop-up to Unmark Everybody in File and Mark Everybody in List

If all the cards don't have the source no, you will have to find and mark them one at a time.

2. Identify the ancestors of the source person who are in the Domesday list -

Go to the Source person's family card then select Options/Direct Lines Bold.(or command-D) for him/her checking the 'Make All Child Buttons Plain' in Step 1.

Find Anything
Attributes/Person Marked/Yes
AND
# of Bold Chldren Buttons/More Than/0
AND
# of Parent Cards/Equal/0

In the list that appears, use the Mark pop-up to Unmark Everybody in File and Mark Everybody in List

3. Mark all the descendants of the Domesday people who are ancestors of the source-

Toggle through the marked cards. For each one, Find/Mark Groups (command-M) and Mark Descendants of each making sure that the Unmark Everybody First box is NOT checked

4. List the the descendants of the Domesday People who are also direct ancestors of the source -

Find Anything
Attributes/Person Marked/Yes
AND
# of Bold Children Buttons/More Than/0

I haven't tried this but it is probably close!

Frank

Gregg
16 November 2006, 08:19 AM
I think Mary Arthur has the simplest solution for you. To add one thing to what she said -- you can use Find Anything to get a list of marked people who have a relationship. These are the people you will need to get a list of descendants for.

Person - Marked - YES
AND
Relationship - Is Not - [leave empty]

Tim Powys-Lybbe
16 November 2006, 06:24 PM
There is a bit of a contradiction there - to find unrelated persons who are also ancestors (if a desc of them is an ancestor, then they are too). If they ARE ancestors, then they are not unrelated.

Are a husband and wife related? Related means that husnabd and wife share the same blood or genes. Unless they have an ancestor in common (Adam apart) they are not normally said to be related. So this is why the 200 odd Domesday people are not related to one another even though they may all have the same source person on the database as their 30th (or so) great-grandchild.

But my thought for Find Anything is -

1. From your source person, do Direct Lines Bold (first unbolding everyone) for ancestors of the source to thus identify the source's ancestors.

2. Somehow Mark only everyone on this list of unrelated persons.

Then with Find Anything, #Bold children >0 (ie an ancestor) AND Person = Marked.

Thanks.

Step 2 is either easy or it is very difficult. Marking the 200 odd Domesday people is easy. Marking the ancestors of the source person who are also descendants of the Domesday people is not easy, it requires lots of going round the circle, which is why I was aking if I could use the Mac Automator.

Doing the query on the marked Domesday people alone just gives you the same people.

Tim Powys-Lybbe
16 November 2006, 06:25 PM
Tim, I am not understanding what you want to do. If family members in your database are connected to one another, then all ancestors and connected people should be able to be found. What are you searching for, exactly?

Martha

I tried to explain what I was trying to do in yesterday's post. Did this make sense for you?

Tim Powys-Lybbe
16 November 2006, 06:54 PM
You may be able to do this in four steps...
Thanks.


1. Mark all the Domesday people -

Assuming the source no xx that you have assigned to the Domesday Book source appears somewhere on all their family cards...

Go to List/Sources (command-S)
Select the Domesday Book source no xx
In the Source Tools pop-up select Show Links to Source no xx (command-L)
Use the Mark pop-up to Unmark Everybody in File and Mark Everybody in List

If all the cards don't have the source no, you will have to find and mark them one at a time.


Unfortunately the source item is a book and it is also used as a reference for other people who are in it but are not ancestors. But I have previously saved the list of the Domesday ancestors so this is not a problem.

2. Identify the ancestors of the source person who are in the Domesday list -

Go to the Source person's family card then select Options/Direct Lines Bold.(or command-D) for him/her checking the 'Make All Child Buttons Plain' in Step 1.

Find Anything
Attributes/Person Marked/Yes
AND
# of Bold Chldren Buttons/More Than/0
AND
# of Parent Cards/Equal/0

In the list that appears, use the Mark pop-up to Unmark Everybody in File and Mark Everybody in List
This produces a list of Domesday ancestors who have no ancestors, it excludes those of the Domesday people who are ancestors of the source person. Fortunately this step does not really seem to be necessary, I can just start with the list of 200 odd marked ancestors.
3. Mark all the descendants of the Domesday people who are ancestors of the source-

Toggle through the marked cards. For each one, Find/Mark Groups (command-M) and Mark Descendants of each making sure that the Unmark Everybody First box is NOT checked
This is where it gets tricky. We start off with 200 marked people. We mark the descendants of one of them, possibly adding in a 1,000 people. How do we know which is the next person in the list to find their descendants? Reunion does not have multiple marks, just the one. The answer is to use a flag to indicate the original list, and unset the flag as each person is done. Further you probably also need to set a flag for all the descendants as each new lot is added in. This gets very complicated to do without making any mistakes and is why I asked if the Mac Automator could be used.
4. List the the descendants of the Domesday People who are also direct ancestors of the source -

Find Anything
Attributes/Person Marked/Yes
AND
# of Bold Children Buttons/More Than/0

I haven't tried this but it is probably close!

Frank

Yes Step 4 will certain get the right answer. But it is step 3 that is the one I could never complete satisfactorily; I know I make mistakes in doing complicated iterative processes like that.

Thanks for thinking about the problem.

Tim Powys-Lybbe
16 November 2006, 07:23 PM
I think Mary Arthur has the simplest solution for you. To add one thing to what she said -- you can use Find Anything to get a list of marked people who have a relationship. These are the people you will need to get a list of descendants for.

Person - Marked - YES
AND
Relationship - Is Not - [leave empty]
Thanks. Let's put Mary's suggestion in here:
I think I understand what you want, first mark the 200 people, unmarking all the others in the data base, and then go to the card of the person you want to know about and then calculate her ancestors (List --> Relatives) and then see which of the people in the list is marked. You will still have to follow each 'trail' separately, but this might narrow it down?
After Mary's suggestion we have the 200 odd marked Domesday People and we also have all the relations of the source person with their relationships. Some of these relations will be marked, from her first step. We then run Gregg's query and it lists the original 200 odd people, nothing more, nothing less. We have not made any progress.

The only next step is to do the procedure that Frank Mitchell suggested but this will not solve my personal problem of getting lost half way through a very long procedure. That is why this thread is titled "Can I use Automator with Reunion". The answer to this question is NO. So I have produced another question: is there anyway of doing this in one (or two or three but NOT a Big Lot!) Find Anything query. So far the answer to this seems to be NO as well.

This is looking like one of those mathematical puzzles that takes decades to resolve and earns the solver a Nobel Prize. (Or is it plain impossible with this release of Reunion?)

Thanks to all for offering so many helping hands.

martha
18 November 2006, 01:59 AM
I tried to explain what I was trying to do in yesterday's post. Did this make sense for you?

Yes, thanks, Tim. I wrote that before you posted your explanation.

Martha

ttl
18 November 2006, 11:47 AM
This is looking like one of those mathematical puzzles that takes decades to resolve and earns the solver a Nobel Prize. (Or is it plain impossible with this release of Reunion?)If I could try a shot at this conundrum, I envision a 2 phase approach.

Phase I:

Logically, if a Domesdayer is an ancestor of the source person's ancestor, the Domesdayer is an ancestor of the source person. Domesdayers who are not ancestors of the source person are not ancestors of the source person's ancestors and thus are of no interest. This phase identifies all Domesdayers who are ancestors of the source person.

- Mark or flag all Domesdayers (all others unmarked)
- Direct line bold all the source person's ancestors
- Do a find anything on all who are marked/flagged AND are bold

This finds all Domesdayers who are ancestors of the source person. The hope here is that this is a manageable subset of the total 200 Domesday set. These can be marked (unmarking first). This list might also be printed out. You can also save this set so you can easily retrieve it again in the future.

Phase II:

With the winnowed set of Domesdayers, go one by one and mark their descendants, checking each ancestral Domesdayer off your printed list as you complete their marking. DO NOT use the unmark first option... you need to allow all the marked descendants to accumulate from one Domesdayer to the next. If you have a set of 30 Domesdayers, this shouldn't be too onerous; if you have over 100, I'd be looking for Automator too. (This is what makes or breaks this approach.)

Once all the descendants of the ancestral Domesdayers are marked, do a find anything for people who are marked AND are bold. This is the list of people I think you're looking for.

Tim Powys-Lybbe
21 November 2006, 06:52 PM
If I could try a shot at this conundrum, I envision a 2 phase approach.

Phase I:

Logically, if a Domesdayer is an ancestor of the source person's ancestor, the Domesdayer is an ancestor of the source person. Domesdayers who are not ancestors of the source person are not ancestors of the source person's ancestors and thus are of no interest. This phase identifies all Domesdayers who are ancestors of the source person.

- Mark or flag all Domesdayers (all others unmarked)
- Direct line bold all the source person's ancestors
- Do a find anything on all who are marked/flagged AND are bold

This finds all Domesdayers who are ancestors of the source person. The hope here is that this is a manageable subset of the total 200 Domesday set. These can be marked (unmarking first). This list might also be printed out. You can also save this set so you can easily retrieve it again in the future.
Thanks for looking at this. This first stage is easy and comes up with 157 people.

Phase II:

With the winnowed set of Domesdayers, go one by one and mark their descendants, checking each ancestral Domesdayer off your printed list as you complete their marking. DO NOT use the unmark first option... you need to allow all the marked descendants to accumulate from one Domesdayer to the next. If you have a set of 30 Domesdayers, this shouldn't be too onerous; if you have over 100, I'd be looking for Automator too. (This is what makes or breaks this approach.)

The number is currently 157, with one more found yesterday. And this is the problem, the query cannot be done solely within Reunion, it needs a printed list, a pen and requires loads of significant iterations. After each iteration, I would want to set a flag for all the newly marked people, thus making it more complicated but a little more secure. I agree that this would work but it does not meet the criterion of doing the search with no more than three Find Anythings; I would get lost or start to make mistake before I had got more than half way through the 157 people.
Once all the descendants of the ancestral Domesdayers are marked, do a find anything for people who are marked AND are bold. This is the list of people I think you're looking for.
If only there were an easy way of doing this! Thanks for thinking about it.