Species missing from Site Details

[IanMiddlebrook]

Hi @DavidRoy,

A recorder has noted that Small Pearl-bordered Fritillary is missing from the list of species seen on her transect, when she looks at the Site Details page: https://ukbms.org/site-details-map?location_id=245429 Yet it was clearly recorded, according to the Annual Summary page and other Reports.

I doubt if this is the only discrepancy. How is the data for Site Details page produced?

Thanks, Ian

[DavidRoy]

@burkmarr Can you investigate this page you developed for the UKBMS. Looks like a species dictionary issue and not picking up the right names?

[burkmarr]

The report applies a filter to the records for a site to ensure that it retrieves only butterflies. For some reason this filter was very complex:

 select count(*) > 0 
 from (select unnest(cof.taxon_path) taxon_meaning_id) taxon_path 
 join (select taxon_meaning_id from cache_taxa_taxon_lists
 where taxon in ('Hesperiidae', 'Lycaenidae', 'Nymphalidae', 'Papilionidae', 'Pieridae', 'Riodinidae')
 and taxon_list_id=15) bfamily on taxon_path.taxon_meaning_id=bfamily.taxon_meaning_id

I don't know why I created the filter like that when this one-liner should do the job:

cof.taxon_group_id = 104

So I can update the report accordingly and Small Pearl-bordered Fritillary is included.

However, it begs the question why the more complex filter removes this species? The reason is that, for some reason, the taxon_path field in the cache_samples_functional table is not populated for Small Pearl-bordered Fritillary records. @johnvanbreda - I suppose that might affect other reports.

[burkmarr]

I've amended report and will post here when implemented on live site.

[IanMiddlebrook]

Hi, Whilst this issue is open, I've had more feedback from a volunteer. Firstly that Dingy Skipper and Chalkhill Blue were missing from her species list (which I would guess is already addressed by the change made?). But also that she would expect to see the species listed in taxonomic order - they seem to be listed in alphabetical order of scientific name. Most recorders get used to seeing species listed in taxonomic order, so this would be a welcome improvement.

Many thanks, Ian

[IanMiddlebrook]

Hi @DavidRoy @burkmarr I'm not sure if this is still a work in progress, but I've had a different comment back from a recorder saying that their site list includes a whole array of unusual species that have never been recorded there: https://ukbms.org/site-details-map?location_id=193315 I investigated, and discovered that, for some reason, they had entered a series of zero counts for these extra species on one date (5/9/2019). Is it possible to exclude zero counts when calculating the species (and number of records) for each site?

Thanks, Ian

[burkmarr]

I updated reports for this but they haven't been pulled through to the live site yet. I've scheduled some time to revisit them early next week and I will look to add the filter for negative records in.

[burkmarr]

I've added a filter to this report to remove any marked as zero_abundance. Also there was a major problem with the SQL which led to some species records being double or triple counted (due to a lack of filter for preferred sp name on joined taxa table) and this has been fixed. I've pulled into master branch and hope to pull into live warehouse today.

[burkmarr]

These fixes are pulled into the live Warehouse now. @IanMiddlebrook - does the Lin Park site look better now? https://ukbms.org/site-details-map?location_id=193315 (I didn't look at it before applying the fix.)

[IanMiddlebrook]

Thanks @burkmarr
Linn Park looks sensible now, an all species are present on other sites.