Filter Question

[wunner 0]

I've tried and tried, but I can't figure out a filter that does the following: I've acquired a lot of cemetery interment lists, and while many include both death dates and burial dates, some only report burial dates. As a result, when I have a burial date of, for example, "17 Oct 1904," I've entered "Oct 1904" as the death date...without recording a specific day. With so many death indexes (indices?) now available online, I'd like to start looking for the exact death dates of these people. Because I probably have a few hundred entries (out of >10,000) that would fit this situation, I tried to create a filter that has a full burial date but only a partial death date, missing only the specific day (as I've noted above), but I haven't been able to figure out a filter that works. Am I trying to do something that can't be done?

 

John

[Michael Hannah 0]

What may not be obvious is that the "Comes before" operator is based on characters not numeric values.

Try a List of Events with a filter of:

 

Tag Type Group is Death AND

Date - Day

 

In the second line you simply put the single digit one as the value.

 

I think this will get what you want.

 

=====edit=====

I just re-read your question, and this will only get you close, but depending upon the number it may be good enough. To refine it more I would have the List of Events set a temporary Flag to 'Y' as a Secondary Output. Then run a List of Events with a filter of:

 

Tag Type Group is Burial AND

Date Is Empty END

 

to set the same temporary Flag to 'N' to exclude the people who don't have a Burial date.

 

I think this should get closer to only the people you want.

Edited May 12, 2008 by mjh

[Helmut Leininger 0]

What may not be obvious is that the "Comes before" operator is based on characters not numeric values.

Try a List of Events with a filter of:

 

Tag Type Group is Death AND

Date - Day < Comes before 1 END

 

In the second line you simply put the single digit one as the value.

 

I think this will get what you want.

 

=====edit=====

I just re-read your question, and this will only get you close, but depending upon the number it may be good enough. To refine it more I would have the List of Events set a temporary Flag to 'Y' as a Secondary Output. Then run a List of Events with a filter of:

 

Tag Type Group is Burial AND

Date Is Empty END

 

to set the same temporary Flag to 'N' to exclude the people who don't have a Burial date.

 

I think this should get closer to only the people you want.

 

Hi John,

 

I think Michael's proposal may not give the results you want because you *have* dates that are incomplete. Michael's solution works for *no* dates.

 

In my opinion, you may try the following (I did not try it myself yet):

 

Create a List of Events for the element-group "death" and filter for date-day before 1 (Michael's trick)

 

Regards

Helmut

[Helmut Leininger 0]

Sorry, I just see that my solution is the same as Michael's. I was too fixed on his second part with date empty and creating an additional outpurt.

 

You should get the same result as in Michael's second suggestion with:

 

Tag Type Group is Death AND

Date - Day < Comes before 1 AND

Date Is Not Empty END

 

Regards

Helmut

[wunner 0]

Michael and Helmut: Thanks for the replies. But from your answers, it seems that you too are finding that a solution to my likely unique problem is not that obvious...that's why I've been struggling.

 

Maybe the solution requires two steps? First to capture all people who have a full burial date, and then a second filter for those people who do not have a "day" of death (just a month and year). I know how to do it when I have a full burial date and the death date is blank, but not when I have a full burial date and only the month and year of death. Now if only I knew how to filter a filtered group!!!

 

John

[Helmut Leininger 0]

Michael and Helmut: Thanks for the replies. But from your answers, it seems that you too are finding that a solution to my likely unique problem is not that obvious...that's why I've been struggling.

 

Maybe the solution requires two steps? First to capture all people who have a full burial date, and then a second filter for those people who do not have a "day" of death (just a month and year). I know how to do it when I have a full burial date and the death date is blank, but not when I have a full burial date and only the month and year of death. Now if only I knew how to filter a filtered group!!!

 

John

John,

 

I think if you specify date - Day before (less than) 1 then it should not take full dates (the day here should be 1 - 31) - at leqast it did not in my trial. So probably only one run should be enough.

 

Regards

Helmut

[Michael Hannah 0]

Maybe the solution requires two steps? First to capture all people who have a full burial date, and then a second filter for those people who do not have a "day" of death (just a month and year). I know how to do it when I have a full burial date and the death date is blank, but not when I have a full burial date and only the month and year of death. Now if only I knew how to filter a filtered group!!!

John,

 

First, Helmut is correct, my single filter finds dates with no day but also includes dates that are completely empty. His added filter to be sure the date is not empty is appropriate for your purposes:

Tag Type Group is Death AND

Date - Day

Date Is Not Empty END

 

Second, I am sorry if my edited answer was unclear (probably too brief) about how to filter this filtered group to be sure you also have a burial date for these people that you have identified with a death date but no day. The way I filter a filtered group is the method I (briefly) explained using a flag. To describe it in more detail, it is really three steps:

 

1) Create a temporary flag where everyone in the dataset is set to the same value, e.g. 'N'. This is easy since simply creating a new custom flag will automatically set everyone to have the defined default value of that flag.

2) Run a report with the first filter to change the flag using the Secondary Output to some other value for those people that match the filter, e.g. 'Y'.

3) Run a second report with the second filter to identify those people that you do not want to be in the first group, and set the flag using the Secondary Output to something different than what you set in the first report, e.g. back to 'N'.

 

It is likely that this second filter may be changing flags on some people that did not have their flag set via the first filter, but that doesn't really matter because it does change those that did have their flag set that you now wish to exclude. The key point is that a Secondary Output flag action does not change the flag value for anyone that does not match the filter used on that report. So if they don't match the second filter, the 'Y' people will stay 'Y'.

 

In step 2) I would use the filter to set the flag for the people that have a death date but the day is blank. Then in step 3) I would use the filter to remove from that group those people who have no burial date. The people that are left with the flag set from the first filter are those that you want, in other words they are filtered from the filtered group. Then based on this flag you can either filter the Project Explorer or create a Focus Group to only look at these people. Once I am done with the flag I usually delete it.

 

Hope this better explains my method for filtering a filtered group,

[wunner 0]

Thanks Michael for a thorough walk-through. The good news is that it works!!! (I never thought of using flags in that way.)

 

John

[Michael Hannah 0]

Great, John,

 

Glad it got you what you wanted. As for flags, this was one of the (many) reasons that custom flags and Secondary Output were introduced.