Show Project Statistics

Post by **tatewise** » 08 Nov 2012 11:28

In response to Wish List Ref 501 File Summary Display I have written the Show Project Statistics V1.0 Plugin now available from the Plugin Store.

Feedback on what other summary Project statistics might be of interest would be welcome.

I intended to include the number of Individuals in each Relationship Pool, but fhCallBuiltInFunction('RelationPool',ptrRec) is not currently working.
This has been acknowledged by Calico Pie and will be fixed in the next minor FH Version.

ID:6575

BillH · Post by **BillH** » 08 Nov 2012 17:40

Mike,

I gave it a try. Very nice. I like the stats displayed.

I do have a couple of oddities though. Here is a screen print of my facts tab:

Note that for Marriage the min, av, and max all say 33. I actually have a lot of different marriage ages, some younger and some older than 33.

Note that for Death it has a min of 9 and a max of 30. I have deaths younger than 9 and older than 30.

Thanks,

Bill

johnmorrisoniom · Post by **johnmorrisoniom** » 08 Nov 2012 17:41

Excellent plugin (as usual) Mike.

My first run threw up an unusual fact of census max age = 444.

Obviously I have had an input error somewhere, but how would I find it?

johnmorrisoniom · Post by **johnmorrisoniom** » 08 Nov 2012 17:45

Decided to include a screenshot, but you can't edit a post to include one

Post by **tatewise** » 08 Nov 2012 20:35

Bill, can you confirm that for the Age anomalies you quote, that the Age fields are actually filled in with a value.
The Plugin currently only looks at actual Fact Age field values, NOT what FH estimates the Age to be, although that might be a useful enhancement.

John, make a Custom copy of the Standard All Facts Query, and add a Column for Age %FACT.AGE%.
Then sort the Result Set on the Age Column.

johnmorrisoniom · Post by **johnmorrisoniom** » 08 Nov 2012 21:51

Thanks for that Mike,
I created the query, but there are no census ages of 444, highest is actually 100.
Also looking at the births entry there is a range of 23-84 when all should be zero, or am I interpreting it incorrectly.

Post by **tatewise** » 08 Nov 2012 22:22

John, can you try adding Row filter Add if %FACT.AGE% is older than 400.
It is possible that the Age is recorded as > 444 yrs or < 444 yrs that sort alphabetically near the beginning of the Age Column, rather than the end.

I agree the Birth Ages look odd, but please try using Row filters:
Add if =FactLabel(%FACT%) matches Birth
Exclude if %FACT.AGE% is null
to check that all Births have no Age values.

BillH · Post by **BillH** » 08 Nov 2012 22:24

Mike,

That is exactly what is going on.

I only have two ages recorded for deaths out of a total of 2988 deaths. One is 9 and one is 30. So the plugin is reporting the correct values. For me at least, having the 'age at death' also reported would be a great enhancement. Same thing goes for my marriage facts. Only one has an age recorded and it is 33.

For all facts I guess it would be great to see the 'age at ...' reported in addition to the actual recorded date.

Thanks,

Bill

johnmorrisoniom · Post by **johnmorrisoniom** » 08 Nov 2012 23:38

Thanks for the help Mike.
Had one lady recorded in 1851 as 444 instead of 44 and 1 in 1881 recorded as 226 instead of 26.
The birth entries occurred when I had V4 and the cursor used to go to the top line when you switched away and back, so the ages were actually the age at death.

results make more sense now as:

Post by **tatewise** » 08 Nov 2012 23:52

WOW!!!
John, I had not expected a statistics Plugin would detect such data entry anomalies.

Bill (and John), for the Age At statistics, should ALL Facts with a Date be included, or only those without an explicit Age field?
Other than that, I think I have coded the Age At statistics.

BillH · Post by **BillH** » 09 Nov 2012 00:22

Mike,

For me it would be nice if the 'age at' was included for all facts that you decide to include on the facts tab. I rarely enter an age when creating a fact. I do this only if I don't know the actual date of birth or the actual date of the fact. If I know both of these dates, I enter them and let FH determine the age.

Bill

TimTreeby · Post by **TimTreeby** » 09 Nov 2012 09:18

Mike,
Have used this plugin and think it is great.
I think for Age Stats, all Facts with a date should be included, but use actual Age if given, if not then use what FH estimates the age to be.
Main reason for this is that most of those who died before 1813 or who married before 1837 their stats won't get included as normally no explicit age is given on either the burial record or the marriage record.

John,
Hope you have noticed that you have a Christening taking place way into the future.

johnmorrisoniom · Post by **johnmorrisoniom** » 09 Nov 2012 12:02

Thanks Tim,
Well spotted (It should have been 1858)

Post by **tatewise** » 09 Nov 2012 13:28

The next version will highlight future Dates and excessive Ages in Red.
The screen-shot also shows the V1.1 Facts tab with new columns.
Number becomes Count but same Fact count value.
No Sources becomes Source and counts Facts WITH at least one Source citation.
Age counts Facts with an explicit Age field, followed by Min., Ave. & Max. for those Ages.
Age@ counts Individuals with an Age At for the Fact, followed by Min@, Ave@ & Max@ for those Ages.
Date counts Facts with an explicit Date field.
I also hope to add a row for All Other Facts.

johnmorrisoniom · Post by **johnmorrisoniom** » 09 Nov 2012 15:00

Hi Mike,
I did actually find the 'No Source' Value useful, as it highlighted the fact that I have unsourced items in my file that need revisiting.
maybe there could be both?

Post by **tatewise** » 09 Nov 2012 16:01

Source does the same job. It is simply the inverse count. It is NOT a count of all Source Citations.
If Count = Source then every Fact has at least one Source, otherwise the remainder have No Source.
i.e. No Source = Count - Source

All the other new counters such as Age, Age@ and Date are counts of things that exist.
By analogy, if there is a difference between the Fact Count and these other counts, then something is missing.
For consistency, and to allow the column width to shrink, I swapped No Source for Source.

The exception to the above rule is that the Marriage and Divorce counts for Age@ can be up to twice the Fact Count because there are up to two Spouses per Family Fact.

This will all be explained in the Help & Advice with V1.1.

Post by **tatewise** » 13 Nov 2012 09:43

Version 1.1 of the Show Project Statistics Plugin is now available from the Plugin Store.

Of necessity, it will reset your statistics the first time it is run, and so you will have to regenerate the new statistics.

The Pool counts are NOT working at present due to an acknowledged bug in the FH API.
As soon as it is fixed, in a minor FH update, the counts should appear.

Let me know if you have any problems with these statistics.

BillH · Post by **BillH** » 15 Nov 2012 00:27

Mike,

Version 1.1 looks good. I really like the mix of stats that you are displaying.

I did have an odd one show up though. Under occupation, I have one person showing as being -11.

I ran a query and found that this person has two birth dates listed, Oct 1869 and 3 Jan 1882. The occupation date is 1 Jun 1870. The occupation was on a census and there is also a census fact for 1 Jun 1870.

I noticed on the Facts tab in the property box that FH is showing the age at occupation as being -12, but the age at census is showing as 8 months. The 8 months came from Ancestral Sources and is in the actual age field for the census fact. It is interesting that FH is coming up with -12 for the age at occupation while the plugin is coming up with -11 for the age at occupation.

Do you know if there is a way to get FH and the plugin to use the earlier birth date in the calculations?

Thanks,

Bill

Post by **tatewise** » 15 Nov 2012 10:05

I have performed a few checks with multiple Birth Events.
FH always uses the first %INDI.BIRT[1]% Event as the reference in its Facts tab Age column, and the AgeAt() function used by the Plugin.

This implies that your %INDI.BIRT[1]% Event is 3 Jan 1882, and your %INDI.BIRT[2]% Event is Oct 1869?.
i.e. They are in the wrong order and need Tools > Re-order Out-of-Sequence Data.

It appears that the AgeAt() function incorrectly calculates negative ages.
I will do a bit more analysis and report it to Calico Pie if necessary.

johnmorrisoniom · Post by **johnmorrisoniom** » 15 Nov 2012 11:25

Hi Bill
Moving away from the plugin.
Your data would suggest to me that you actually should have two individuals.
Child A bn 1869 and died sometime after 1870.
Child b bn 1882

It is/was quite common that when a child died, the next child of the same sex born would re-ude the name.
I have 4 consecutive Robert Henry's, the first 3 only lasted a few months, the last one a few years

BillH · Post by **BillH** » 15 Nov 2012 17:11

Mike,

Don't I feel stupid. I should have thought of that. [oops]

Thanks!

Bill

BillH · Post by **BillH** » 15 Nov 2012 17:14

Hi John,

Very good point. I actually have a note on the person saying that this might be two different people. I guess I was waiting to find more evidence before actually splitting out another child in FH, but I think I'll go ahead and do it now.

Thanks!

Bill

Post by **tatewise** » 15 Nov 2012 21:02

Bill, I am amazed at how many database anomalies these simple statistics are unintentionally revealing.
They seem to highlight various abnormal data patterns that warrant further investigation.
I don't claim that all abnormalities will be detected, but it all helps.

Could those of you with large databases give me some idea of how long the Plugin takes to run.

RogerF · Post by **RogerF** » 16 Nov 2012 15:36

Mike said

Could those of you with large databases give me some idea of how long the Plugin takes to run.

14000 individuals in 19 seconds.

johnmorrisoniom · Post by **johnmorrisoniom** » 16 Nov 2012 17:12

Hi Mike
V1.1
Indiv: 32543
families: 8871

Run Time just under 2:30

Family Historian User Group