* Sorting apparently duplicate facts

[mardler]

Posted a question on this topic in the wish list section earlier and didn't realize it probably doesn't get much viewing. So here goes again.

I'm a newbie getting used to FH having previously used FTM Some of the subtleties in approach between the two programs are still emerging for me.

I have lots of what might at first sight be duplicate records, eg, death dates, burial dates, birth dates and names. (It's a result of my being a member of the Guild of One-Name Studies, needing to record every instance of a single name and wanting to show my sources for people to check my work.)

FTM had a useful little button allowing me to specify which of a series of similar facts is preferred. FH doesn't appear to have a similar button for anything other than names, but now I see that the main fact appears to be the first in the list, ie, the oldest/earliest occurrence of a particular fact. Am I right?

For example, if I have birth facts for an individual as
1864, c1865, 15 Apr 1865 and 1866,
FH shows the date of birth as 1864 when I want it to show 15 Apr 1865.

Am I correct in my assumption? Am I doing something wrong? Is there a fix?

Perhaps I should add that I use TNG for web display and TNG uses the preferred facts as the primary dates for indexing individuals. (Since I have 1262 people named William xxx HOWES alone, being able to find individuals from a list of exact dates where known is a help!)

TIA for any help
Paul

ID:7060

[tatewise]

Essentially FH lists all Facts in Date order.
There is even Tools > Re-order Out-of-Sequence Data to help ensure that is the case.

If a Fact has no Date, then its Normal Time Frame is used to determine its position in the list, e.g. Post-Death Facts will appear after any Life Facts.

This is the usual order of the Facts listed on the Facts tab of the Property Box.

The Main tab shows a summary of some Facts, and by default the earliest of any multiple Facts are shown.
However, the Property Box can be customised to show multiple Facts, either on the Main tab or on a custom tab that you define.

Basically, all your assumptions are correct, and FH does not have a preferred Fact option.

However, you could create Custom Facts to hold your preferred data.
e.g. a Primary Birth Fact would hold 15 Apr 1865 Birth details, and similarly for Marriage and Death.
Then the Main tab could be customised to show these Primary Facts.

There are several other ways you could identify one of the Standard Facts as the Preferred Fact (e.g. a unique Note), and list them in the Records Window, in a Custom Query, or using a Plugin.

[mardler]

Thanks for the super quick reply, Mike.

Something surprising is happening, though. I am beginning to use FH for data entry, rather than just transfer data across from FTM. Let me relate what I just experienced.

I had only the barest of data for one person: a name and a year of birth. I just added birth registration for that person, adding a source for the name and a new birth fact which included the month and the place, ie, 'c Nov 1900' and 'Mitford RD, Norfolk'.

The result of doing that was that the second birth fact is now referred to as 'Birth*' by FH. The original fact has apparently been relegated to 'Birth[2]*'. This is a perfect outcome for me, at least in a simple case.

Is that clever programming? In most family history programs, '1900' would appear ahead of 'Nov 1900' in a chronological list, but in FH they are reversed.

Regards and TIA
Paul

[tatewise]

I have investigated the way FH sorts Dates and discovered some subtleties I was not aware of.

It seems that the rules FH uses when you click on the Date header or use Tools > Re-order Out-of-Sequence Data are as follows:

1) A Fact Date in an invalid position is moved up or down as little as possible until in a valid position.

2) Any Fact Date in a valid position will retain its All tab Record order.

What determines a valid Date position is interesting.
Firstly, Period, Range & Quarter Dates use the earlier Date.
If the natural interpretation of the Fact Date is OK then it is in a valid position.
e.g.
A Fact with no date can be before or after any other Date as long as its Normal Time Frame is OK.
1980 can be before of after any other Date in the same year.
May 1980 can be before or after any other Date in the same month.

So in your example, you can use the All tab to put 1900 and Nov 1900 in either order.
In your original posting 1864 must be first, c1865 & 15 Apr 1865 can be either order, and 1866 must be last.

Note that the asterisk * against the Fact in the Sources pane, simply indicates that there is Source Citation.

[mardler]

Thanks again, Mike. It was flexible programming!

I'll fiddle with the records I can, but it sure would be handy to be able to adjust those fact suffixes and designate a preferred fact. As I say that, I know I'm looking backward rather than forward and I'll try to get over it!

Regards
Paul

[tatewise]

Those 'fact suffixes' are the positional indices that are adjusted by sorting and moving, and are part of the GEDCOM specification that FH directly supports.
The 1st is [1] but often not shown, the 2nd [2], the 3rd [3], ad infinitum.

Here is a way of implementing a preferred Fact strategy.
It relies on the Fact sub-tag TYPE that FH refers to as Descriptor, and customised columns in the Records Window.
(This sub-tag is rarely if ever used, so is a good candidate, but there are other options.)

Step 1:
Where there are multiple instances of a Fact such as Birth, open the All tab.
Right-click on the preferred Fact, select Descriptor and enter Prefer.

Step 2:
In the Records Window use Lists > Configure Records Window Columns.
Scroll to bottom of Item list, select and click > button to Add Column.
Enter Heading such as Birth Preferred.
Enter Expression such as =ItemIf( %INDI.BIRT[2].TYPE% = 'Prefer', %INDI.BIRT[2]%, ItemIf( %INDI.BIRT[3].TYPE% = 'Prefer', %INDI.BIRT[3]%, %INDI.BIRT[1]% ) )
Click Add and use up/down arrows to position this new Column, then click OK.

This will display the Birth Details of the Preferred Birth, with 1st Birth shown by default.
An explanation of the Expression follows.
The Function used is ItemIf( Condition, Item-True, Item-False ).
Where if the Condition is true, then Item-True is shown, else Item-False is shown.
Note that either Item-True or Item-False can be another nested Function.
=ItemIf(
   %INDI.BIRT[2].TYPE% = 'Prefer' ,   1st Condition
   %INDI.BIRT[2]% ,                           1st Item-True  
   ItemIf(                                          1st Item-False
       %INDI.BIRT[3].TYPE% = 'Prefer' , 2nd Condition
       %INDI.BIRT[3]% ,                         2nd Item-True
       %INDI.BIRT[1]%                           2nd Item-False
   )
)
This supports three Facts, but with further nesting, several more can be supported.

[mardler]

Thanks - step 1 looks simple enough
Paul