Wednesday, February 08, 2012
 
Birder's Diary Forums
Minimize
Subject: ABA Report for Transcription Bug
Prev Next
You are not authorized to post a reply.

Author Messages
PeteThayer User is Offline Posts:101 Veteran Member
01/17/2010 5:23 PM Alert 

Hi Jeff,

Minor Bug Found:
I updated my ABA Report using the link in the Bug Fix section. Everything on the report seemed to work fine - except that the Millenium Total (ML in the Other column) showed a number less than 500, even though I ran the report and said to show only counts over the ABA thresholds (I also said to follow ABA rules). You might have Steve O. look into this.

I have another question for you:
My Observer profile is set to show World counts (Clements 2008 v6.3.2) and ABA totals (ABA 2009). I used exactly the same taxonomic settings when I ran the ABA Report.

But my World total says 3987 at the bottom of the main screen of BD 3.6.107 and the ABA Report I ran shows WO = 3984 when I follow ABA reporting rules and 3987 when I ignore them.

Also, the ABA total at the bottom of my BD screen shows 659 but the ABA Report I ran shows AA = 660 (both when I follow ABA rules and when I do not.)

What would be causing both of these totals to be different? I would like to resolve this before I go to Colombia next month. I have been practicing my 4,000 Lifer dance for some time now, and I do not want to celebrate 3 birds too early and make a fool of myself!!

 


pete@thayerbirding.com
SteveO User is Offline Posts:509 Veteran Member
01/17/2010 5:57 PM Alert 

Hi Pete,

Do any of your other counts loook funny? Is it just World and ABA Area?

Could you run the All Locations report, and compare the counts for the locations that are common between the reports?

One other user has noticed discrepancies in the reports, and I am trying to see if there is a common thread.


Regards,
Steve
PeteThayer User is Offline Posts:101 Veteran Member
01/18/2010 4:56 PM Alert 

When I run the ABA Report WITH thresholds and following ABA rules for species validation I get exactly the same number as when I run the same report WITHOUT thresholds.

When I run the ABA Report WITH thresholds and counting all birds and ignoring ABA rules I get exactly the same number as when I run the same report WITHOUT thresholds.

Differences appear between reports where I follow ABA rules and I do not follow ABA rules. Not following the rules always results in the same or more birds.

AA +3
LF +3
US +4
OU +8
CE +4
WO +3
NA +7
PO +1
TT +3
FL +3
HI +1

And +1 for the Annual List for these locations: LF, AA, TT, OU, FL, NA, US

For the annual list, I see 110 for FL when following ABA rules and this is the same number I see when running a life list report for FL for 2009 using the ABA-2009 taxonomic list. However, when I use the Clements taxonomic list, Purple Swamphen is added and the total goes to 111. Purple Swamphen is not on the ABA official list.

I guess my real questions are:
How does someone determine what the "extra" species are for a location when you do NOT follow ABA rules? What is the SQL query used in BD to generate this? And finally, what is the query being used to compute the totals on the bottom of the Main screen of BD and why is it a different SQL query than that used to create the ABA Report? My ABA totals of 659 vs. 660 are the really strange results in all this!


pete@thayerbirding.com
jjones User is Offline Posts:5092 Veteran Member
01/18/2010 8:12 PM Alert 

Hi Pete,

This has come up before. Everything in BD uses the same algorithm (SQL query and calculations) to calculate life list counts, EXCEPT for the ABA special report. Steve did something different there. You can trust the Counts at the bottom of your main screen for the selected observer, taxonomic list and location.

Steve will address the rest of the above when he gets a chance. I think Steve has a handle on why it is producing different numbers but I will let him respond to that.

thanks,

Jeff


Birder's Diary Technical Support
SteveO User is Offline Posts:509 Veteran Member
01/18/2010 9:18 PM Alert 

How does someone determine what the "extra" species are for a location when you do NOT follow ABA rules? What is the SQL query used in BD to generate this? And finally, what is the query being used to compute the totals on the bottom of the Main screen of BD and why is it a different SQL query than that used to create the ABA Report? My ABA totals of 659 vs. 660 are the really strange results in all this!
Pete - several interesting questions that go back in history to version 2.0. I wanted some reports that were not performing at the time - ABA/All Locations, and Comparison.

The ABA report existed, but took forever to run, and All Locations and Comparison did not exist. When I tried the normal algorithm, a query for each node in the whole location hierarchy using different tax lists did not work out.  In order to code programs that would produce the ABA/All Locations report in less than a fortnight, I used a different processing method. These programs were first offered as an add-on, and with version 3, they were adapted for the new database and became part of the product.

Even with improvements in processing speed, I do not know if a variation of the "normal" query would be able to produce the reports in a reasonable timeframe - but I have not attemped to code it since the existing report seems to work - at least for most people.

At the current time, I am working on a method of showing which taxa have been seen at which location. I have the code in a prototype now, and would like to run it against your database, to see if there is an explanation or if my program is just terminally hosed.

If you will upload your database, I will see if I can run the reports and do some analysis. If it works out, I will share the results with you to see if it explains the anomalies.


Regards,
Steve
SteveO User is Offline Posts:509 Veteran Member
01/18/2010 9:57 PM Alert 

And I missed a couple questions

When I run the ABA Report WITH thresholds and following ABA rules for species validation I get exactly the same number as when I run the same report WITHOUT thresholds.

When I run the ABA Report WITH thresholds and counting all birds and ignoring ABA rules I get exactly the same number as when I run the same report WITHOUT thresholds.

Differences appear between reports where I follow ABA rules and I do not follow ABA rules. Not following the rules always results in the same or more birds.
All the Threshold selection does is suppress reporting counts for any location (state or province) that does not have enough sighings to qualify to be reported to the ABA. For example, for 2009 the Florida Threshold is 250 Species. If you have seen less than 250 species, they do not want you to report your Florida count. The threshold switch will suppress that count so that you will not need to compare it to the threshold before you transcribe it.

The ABA Rule switch is also fairly simple. If the ABA rules are in effect,

  • The ABA tax list is used for all locations in the ABA area (Continental US, Canada, and St Pierre et Miquelon)
  • The AOU list is used for all locations in the AOU area that are not included in the ABA area.
  • The Clements list is used for all locations not matching either of the previous two statements.

If the ABA rules are not in effect, Clements applies everywhere - so Purple Swamphen is good.- as are the Ostriches you see when driving from Tucson to Sierra Vista. I would expect this number to be bigger, because more species are not ruled out.

Still looking forward to seeing your database...  And to seeing your 4000-species victory dance on YouTube.


Regards,
Steve
You are not authorized to post a reply.
Forums > Birder's Diary > Technical Support > ABA Report for Transcription Bug



ActiveForums 3.6
  
Support Tools
Minimize
 TitleOwnerCategoryModified DateSize (Kb) 
  
Copyright 2011 by Jones Technologies LLC  Terms Of Use  Privacy Statement