Bad coordinates in clog.

[Birger]

I think for

Zaireichthys lacustrus

it may have got punched in 14°07’N instead of 14°07’S seems to be the right distance but north. Should be in Lake Malawi not Sudan.

while I am looking

Tetracamphilius pectinatus

is off the coast of Cape town in the ocean, should be in the Congo.

Birger

[Jools]

I think for

Zaireichthys lacustrus

it may have got punched in 14°07’N instead of 14°07’S seems to be the right distance but north. Should be in Lake Malawi not Sudan.

while I am looking

Tetracamphilius pectinatus

is off the coast of Cape town in the ocean, should be in the Congo.

Hi Birger,

These two are mapped due to co-ordinates in the type locality. Do the descriptions mention the type locality co-ords or have you cleverly figured them out? If the later, then we (you, me, @MatsP) should work out what to do here. Up to now, descriptions with co-ords have been entered as occurrences.

Jools

[MatsP]

I split this out from the announcement thread.

The Zaireichthys appears to be correct, but it was using some funny "right-single quotes", so I replaced that with teh regular single-quote - also updating the name to "Z. lacustris" (so link above will be broken!)


The Tetracamphilus pectinatus appear to be an error - unfortunately, not able to figure out what it should be. It's the same error in CoF and Ferraris - and even the original description. Looking at a map, and following the references in the original description, I have corrected it by removing the 4 in the beginning of the coordinates - this is closer to the "Kibanzi" location mentioned than removing the 5 and leaving the 4.

Thanks Birger for spotting these.


--
Mats

[Jools]

Good idea to split. It's amazing how much of this stuff suffers from transcription errors (at source).

Cheers,

Jools

[Birger]

These two are mapped due to co-ordinates in the type locality. Do the descriptions mention the type locality co-ords or have you cleverly figured them out? If the later, then we (you, me, @MatsP) should work out what to do here. Up to now, descriptions with co-ords have been entered as occurrences.

These two were straight out of the original descriptions, I do use COF at times as well though. I have only used coordinates that were supplied so far, they are sometimes messed up so I should pay more attention to them now.

also updating the name to "Z. lacustris"

Thanks for that, my bad.

If I see the odd one wrong is it quite technical, or is it a quick fix that I can do. Attaching to this thread may be better for me?

So far I have avoided looking thru

Synodontis

Birger

[MatsP]

If it's directly wrong in the "Type location" (e.g. I've copied something from another source, such as CoF, and that source is wrong), then it's just a case of fixing it up in the type locality field as a simple edit (e.g. It actually says 45°07'N, when it's supposed to be 45°07'S according to some other source - then just edit "N" to be "S") - or if the value is 8 degrees west, when you know it's supposed to be in Brazil (and thus between 40-80 degrees west, roughly) - if you can find what the correct value should be, by all means, edit away.

However, the way that we ("we=the code on planet catfish") read the coordinates out of the type locality field is tricky, and it's sensitive to "wrong kind of degree sign" or "wrong kind of quote signs" - there are several different varieties of quotes and a "little 'o' used in Spanish, Italian and Portuguese to note numerical gender (or something)" that is often used as a degree sign in some literature (after all, it looks like a degree sign if you didn't know the difference). These sort of things can be tricky to figure out, and since the way the code that parses this works, if it finds something it doesn't recognise, it stops there with what it got - which may be "wrong side of the equator" or "wrong side of the zero meridian". It's very hard to make the code better. So, if it "looks about right", but the "pin" isn't in the right place, then posting in the Cat-eLog data issues is probably the right thing to do.

[Jools]

Hi Birger,

It should be a quick fix if you're just copying from a description. I think it's a difference between using ` and ' in the text or something similar. I think @MatsP can confirm (and he has as I was typing this!).

Jools

[Birger]

Okay, no problem...in my head I thought you were punching them into another area of PC as well for the maps.

Birger

[Jools]

Okay, no problem...in my head I thought you were punching them into another area of PC as well.

Your head is right! They are gleaned from the type locality if present there. Also, for any species, an admin can record an occurrence (e.g.

Synodontis granulosa

occurs in Lake Tanganyika). If the entry for Lake Tanganyika on the database has co-ordinates (another thing an admin can do) then we see a map lower down in the species profile.

Both occurrence and type locality map data are plotted on the newly introduced family and genus maps.

Jools

[MatsP]

I think it's a difference between using ` and ' in the text or something similar.

The really tricky part here is if you are copying from a PDF or web-site, there are several different quotation marks that look ALMOST the same (for example there are at least two double quote marks ", ʺ and ˝ or single quotes: ', ʹ, ʻ, ʼ and ʽ - if you see ONLY one of these, it's hard to spot that it's not "the right one"). My code actually does (try) to cope with several variants, but there are quite a few, and I've certainly found cases where it goes wrong... So if it "looks about right, and the marker is not where you think it should be", I can have a look at the text and see if it's using "strange quotes" and the like. [Note if we have 45˚07`S - it will become 45˚N, because it doesn't recognise ` as a "correct sign for minutes". I've also had problems with strangely formatted strings, such as 2-3˚S, which gets misread...]. So in summary, there are a lot of things that CAN go wrong here - and the devil is in the detail.

It is of course also possible that the coordinates for the body of water, say, Lake Tanganyika, is wrong, and it doesn't show the marker where it is supposed to go - so if you see something that is "wrong" in that respect, then also let us know.

--
Mats

[Acanthicus]

Hi,

N. furcatus is in the middle of the indian ocean: http://www.planetcatfish.com/catelog/ge ... d=275#3795
and
M. draco is on the east coast of Mozambique: http://www.planetcatfish.com/catelog/ge ... d=386#2567

[Jools]

@Mats,

You suggested this may be a re-interpretation problem? Are you looking at it or would you like me to have a look?

Cheers,

Jools

[MatsP]

@Mats,

You suggested this may be a re-interpretation problem? Are you looking at it or would you like me to have a look?

Cheers,

Jools

I've had a look, but I don't understand it. I wonder if some browsers are re-issuing the text in a different form or something like that. It's either that, or I can't replicate the problem because it's some difference between my machine and your hosting service's machine(s).

I think I'll try adding some logging to the vetting code, and then you can send me the file [in the usual bundle of files?].

--
Mats

[Jools]

Need anything more from me Mats?

Cheers,

Jools

[MatsP]

Need anything more from me Mats?

Cheers,

Jools

I'm starting to suspect it's something to do with how different browsers "post" the updated form. I will submit some changes that log the exact content when we parse the type locality.

--
Mats

[MatsP]

Just submitted R1017 (minor comment fixes for constants) and R1018 with the above fixes. It creates a log-file in `/log/vet.txt` - if you could add that to the regular package you send me, I'd be able to keep an eye on what it does.

--
Mats

[Jools]

The logging version is now online.

Cheers,

Jools

[Jools]

This now appears to be fixed (the bug causing co-ordinates to corrupt over time). Birger, can you confirm?

Anything now that isn't right should be a data issue that can be fixed with a data submission or occurrence amendment and so just let us know if something ain't right.

Cheers,

Jools

[MatsP]

*BUMP* Birger, can you confirm?

--
Mats

[Jools]

Just to be clear, I was asking Birger to confirm the African cats locations are OK, not that we've fixed the bug that corrupts co-ords over time.

Jools

[MatsP]

not that we've fixed the bug that corrupts co-ords over time.

Agreed, I think we have hit the "bug that corrupts co-ords" on the head too, but that's a different bug.

--
Mats

[Birger]

I believe Mats has this under control now.

Birger