PlanetCatfish.com

Hi Jools,

For some reason, the CLOG is placing a type locality distribution/occurrence marker for in Saudi Arabia. A similar problem happened before with ; I think that there must be an automated bot that trolls the Type Locality data from Fishbase (or some other source), and perhaps it's swapping N for S and E for W, or it's swapping latitude and longitude... That's just a guess.

But if that is correct, it's not consistent: CLOG updates for both and occurred on the same day. B. minerim shows this mistake, but B. hartii does not.

Oddly, B. hartii shows NO type locality marker on it's Distribution/occurrence map; clearly there are some funny quirks to the distribution maps in terms of how they plot/don't plot these occurrence markers depending on whether or not there are Type Locality Lat/Long data available, and depending on the number of additional occurrences listed/mapped under the Distribution field of data (like here: Faulty Distribution map display for Corydoras schwartzi).

I don't mean to bring this up too often; it's not a big deal and I'm not really concerned about smoothing out the little glitches like the presence/absence of all the flags on a single map, but I imagine that you don't want erroneous markers, as in this case in Saudi Arabia, persisting.

Cheers, Eric

@MatsP really knows how this works and built the original code that parses the lat/long co-ords. Hopefully he can have a quick look at this...

Jools

That said, I checked the type locality and it was 17°44’52''S, 47°05’40''W, I changed it to 17°44'52.0''S 47°05'40.0''W. Note, uses ' and not ’. This was then parsed correctly and the type locality now appears below the equator and west of the meridian. Seems this was the issue?

Jools

Hi Jools,
Thanks for the fix.

Jools wrote:That said, I checked the type locality and it was 17°44’52''S, 47°05’40''W, I changed it to 17°44'52.0''S 47°05'40.0''W. Note, uses ' and not ’... Seems this was the issue?

Although obviously it fixed the problem, I hope you're not asking me, as I don't know how the CLOG obtained the original Lat/Long data in the first place. I can only assume that the CLOG trolls the info from Fishbase or some other source, and that this source must have used the wrong punctuation ( ’ instead of ' ) to begin with. And if this was true, wouldn't the source database make the same punctuation mistake every time, if its database is formatted in a particular way? (unless the data base stores Lat/Long data as character strings rather than numerical values with units of °, ', and '' . In that scenario, I can imagine people using the wrong symbol keys during data entry, and since the info is stored as a character string, there would be no formatting check required to conform the characters used for units).

Cheers, Eric

Sorry, I should have said the clog pulls the type locality data from an online source. Thus no inference of incorrect usage!

Jools

We really should be more generous with what symbols we accept as ' and " symbols. They can be terribly hard to spot the difference (when in doubt, use only the single quote ' - two of them for second, one for minute).

--
Mats

I trolled through each family and looked at the world-wide family-level distribution maps, in search of any species which seemed to be out of place (on wrong continent or in the middle of the ocean when they shouldn't be) because of several errors like the one you've found. I found several different errors:

• ’ instead of '
• ; instead of '
• ' instead of ''

And for one species the coordinates were just plain wrong. Somewhat surprisingly (which was a good thing), I found only five species conspicuously out of place. CLOG corrections have been submitted.

Cheers, Eric