c24m48 writes...

I just purchased, downloaded, and installed Version 2. The first feature I wanted to play with was Partial File Analysis. It seems to have a Descendants option, but not a Descendants and Spouses option. My first wish list item for the next update would be to add a Descendants and Spouses option. However, .....

I put descendants of myself in as a test case, with unlimited generations. I have three grown children. Both daughters are married. My son is not yet married. One daughter has children. One daughter does not. Here's the results.

1. It included me.

2. It did not include my wife (consistent with including Descendants rather than including Descendants and Spouses)

3. It included all of my children (which is correct)

4. It included all of my grandchildren (which is correct)

5. It included one of my sons-in-law but not the other (which is strange)

The son-in-law who was included is the father of my grandchildren. So I was thinking it includes spouses if there are children, and doesn't include spouses if there are not children. But if that is true, then why didn't it include my wife?

Jerry Bryan

On 2007-05-10 8:57pm GenSmarts replied...

Here's the way the rules are written - the intent was to cast a wide net :

All children of the root person are considered descendants

Siblings of children are considered descendants (beyond the first generation beyond the root)

Spouses of descendants are considered descendants as well (beyond the first generation from the root)

Parents of descendants are considered descendants as well (beyond the first generation from the root)

The root person is simply added to his/her list of descendant and not considered a descendant of his/herself.

You are already exporting your file - what rules have you been using in actual practice to do that? That's what I would hope we'd be able to provide support for (either already or after a few tweaks).

On 2007-05-10 10:31pm c24m48 added...

What I have been exporting is root person, descendants of the roots person, and all spouses (including spouses of the root person and spouses of descendants). I have not been exporting parents of spouses.

In my little test example with GenSmarts, here are the differences when I use myself as the root person.

person --------- exported file ----- GenSmarts

myself ----------------- yes -------------- yes

my wife ---------------- yes -------------- no

my three kids ---------- yes -------------- yes

husband of daughter 1 -- yes -------------- yes

husband of daughter 2 -- yes -------------- no

my grandchildren ------- yes -------------- yes

I would like to be able to include my wife and the husband of daughter 2. It appears that my wife is not being included because she is in the first generation, and spouses are included only beyond the first generation. It is not clear why the husband of daughter 2 is not being included. He is a spouse, and he is beyond the first generation. Plus, the husband of daughter 1 was correctly included.

The GenSmarts inclusion of siblings of children is intriguing. If I understand the feature correctly, this would catch half brothers and half sisters of descendants, where the half brothers and half sisters themselves were not descendants. I could see some times where it would be extremely valuable to include them in my research, but other times where I really wouldn't want to include them. I wonder if whether to include them or not could be an option.

I have the same comments about the GenSmarts inclusion of parents of spouses. It's an intriguing feature. I could see some times where it would be extremely valuable to include them, and some times where I really wouldn't want to include them. Again, I wonder if it could be an option.

Jerry Bryan

On 2007-05-11 1:13am GenSmarts replied...

I suspect...

"Husband of dau1" was included because he was the parent of a descendants (e.g. he has a child who is a descendant)- and your "Husband of dau2" is not (he has no children who are descendants).

On 2007-05-15 11:06am c24m48 added...

Here's a little further experiment with analyzing part of a file. Instead of descendants of myself, I selected descendants of my father. That way, the sample size for the experiment is still small, but I am no longer the root person. Selecting my father as the root person, the list now includes my brother and sister and their descendants, but let's focus on how changing the root person to my father changes the list of my own personal descendants.

Some of my questions may be related to exactly how generations are counted. If my father is the root person, one way to count it is that my father is generation 1, I am generation 2, my children are generation 3, and my grandchildren are generation 4. Another way to count it if my father is the root person is that I am the 1st generation from the root, my children are the 2nd generation from the root, and my grandchildren are the 3rd generation from the roots. GenSmarts appears to be counting the latter way rather than the former way.

With my father as the root person, my wife is now included in the partial file analysis. However, she does not appear to be included because she is my spouse. Rather, she appears to be included because she is a "parent of descendants ... (beyond the first generation from the root)" -- namely, she is the parent of my children and my children are the second generation from the root.

When I was the root person, my children were the first generation from the root. My wife was the parent of descendants at the first generation from the root rather than the parent of descendants beyond the first generation from the root. That's why my wife was not included when I was the root person.

To prove this, I temporarily added a second wife for myself (grin). The second wife was not included in the partial file analysis with my father as the root person. Then, I temporarily added a child for this hypothetical second wife and myself. The second wife was now included, as was the child. (I've got to remember to remove the second wife and extra child from my file!)

Finally, I make note of the following. The son-in-law of my younger daughter (newlyweds who do not yet have children) is not a part of the partial file analysis when I am the root, but my new son-in-law is a part of the partial file analysis when my father is the root. So my new son-in-law does not appear as a "parent of descendants", no matter how many generations I go back to establish a root person. But my new son-in-law does appear as a "spouse of a person beyond the first generation from the root" as soon as I establish a root person far enough back to make him "beyond the first generation from the root".

I would recommend either as an option or else "that's just how it works" that when doing a partial file analysis based on descendants that all spouses be included in the partial file, irrespective of whether the spouses have children that are descendants and irrespective of which generation the spouse is from the root (including spouses of the root person). The way it works now, spouses are always included beyond the first generation from the root person. They are included at the first generation from the root person only if they have children who are descended from the root person. And they are not included at all if they are spouses of the root person.

Despite these concerns, I have been using the new partial file analyis feature for descendants very successfully with real data, not just for little thought experiments. It's great! It's very difficult to convey just how nice it is to be able to use GenSmarts against my production genealogy database instead of against extracts.

Jerry Bryan

On 2007-05-15 8:14pm GenSmarts replied...

Sounds reasonable, I'll let you know when we get the change in.

On 2007-06-11 10:14am GenSmarts replied...

From C24M48:

In Version 2.1.1.10 something strange is happening with "Analyze only part of a file - descendants". For example, if I choose descendants of my paternal line great grandfather, I get 320 descendants. It's hard to verify for sure by hand, but that's about the right number. When I go just one generation further back to descendants of my paternal line great great grandfather, I get 41268 descendants. That can't be right. I can look at the GSF file with Notepad to assure myself that I did the selection correctly, and I did.

The number of "descendants" in the latter case is so large, it's hard to see what's going on. But the 41268 figure represents nearly my entire data base, including huge numbers of people who were not even remotely descendants of my paternal line gg grandfather. For example, the 4126 figure includes people as far back as eleventh great grandparents in collateral lines.

Jerry Bryan

On 2007-06-11 10:20am GenSmarts replied...

The original intent was for this to grab anyone in your file that was there because of the ancestor in question. So perhaps calling it "descendants" is a bit misleading, because it could walk linage behind the person you chose, depending on the max generations set.

In any case, we'll put a change into the next update (2.1.1.15 or higher), to make this more straightforward:

Descendants means children plus spouses of the person chosen , cascaded downward and limited to the "number of generations" specified. We'll also include spouses of the person chosen.

On 2007-06-11 2:22pm c24m48 added...

Sounds great. Thanks.

Jerry Bryan

On 2007-09-19 1:01am c24m48 added...

I just upgraded to version 2.1.1.21. Does this version include the new definition of the support for a partial analysis of descendents?

Jerry Bryan

On 2007-09-19 1:07am c24m48 added...

Actually, it looks like I've answered my own question. It appears that the new support is there. I was having a hard time telling for sure because I was choosing partial files that were small anyway. But when I chose a partial file that included more generations, it was easy to see that the new support is there. Much thanks.

Jerry Bryan

Jerry Bryan

---

Tags:  Untagged  

---