topic Re: Bug with adding XY data? in ArcMap Questions

Bug with adding XY data?

HelenGoodchild — Tue, 30 Jan 2018 10:48:42 GMT

We're using ArcGIS 10.5.1 for an undergraduate class in GIS. In that class four separate people had an issue in adding XY data. In all of these instances, it would plot somewhere wrongly. However, if they went back to the original spreadsheet and changed the column headers from LONG/LAT to X/Y it plotted correctly. It wasn't a simple transposition error of them getting the Xs and Ys the wrong way round as I doubled checked each of them fully.

This seems really odd, and I'd be interested if anyone else has had anything similar. I've had nothing but trouble since our most recent upgrade, so am really considering rolling back versions as it doesn't make for a great learning experience at the moment.

Thanks,

Hen

Re: Bug with adding XY data?

DanPatterson_Retired — Tue, 30 Jan 2018 11:32:57 GMT

Similar reportings... one of yours?

LONG is not a good choice for a field name... potential confusion with a numeric data type.

Try Longitude and Latitude, little chance of confusion.

I prefer X and Y in any event because of the prevalence of saying 'Latitude/Longitude'... the source of the switching problem since mathematically we say X/Y and interpret any other terms as being directly associative. Sadly, saying it the 'wrong way' (IMHO) cause no end of confusion in the class.

So ... in short, be explicit in naming the fields and order them in the table as such, or just stick with X and Y so that Longitude/Latitude's ancestry is solidified (lets not get started on Easting/Northing either)

Re: Bug with adding XY data?

HelenGoodchild — Tue, 30 Jan 2018 11:43:04 GMT

Hi Dan,

Thanks for the clarification. I was just abbreviating - some of the students had written in full and it still caused issues, but good to know that the LONG could cause problems if used! I'll be making sure they use X and Y in future, but have also used Eastings/Northings in OSGB in the past (with no issue).

I know what you mean about the XY-YX LAT/LONG confusion - always throws people (including me)

Cheers,

Hen

Re: Bug with adding XY data?

shan_sarkar — Tue, 30 Jan 2018 12:05:40 GMT

Helen,

Is it possible that your students might have used the incorrect field names like fields with spaces or any special characters?

I have faced this issue on many occasions my clients used to share the data via excel and this was the first thing I used to check and have made it a definite thing on my checklist.

You may refer: FAQ: What characters should not be used in ArcGIS for field names and table names?

I hope this helps!

~Shan

Re: Bug with adding XY data?

HelenGoodchild — Tue, 30 Jan 2018 12:15:44 GMT

Hi Shan,

Thanks - I was suspicious as I always thoroughly check the things on that list, but the data did plot, just not in the right place, and not simply transposed - it was compressed into the North Sea, rather than spread across Northern Europe! I would assume that header errors and formatting would prevent them being visible in the XY drop down boxes rather than plot them incorrrectly? I'm going to make them stick to a rigid convention of X / Y in future.

Cheers,

Hen

Re: Bug with adding XY data?

XanderBakker — Tue, 30 Jan 2018 12:20:23 GMT

You mention that there were "four separate people" that had issues with the file. Was it the same file or did each student create its own? How many people managed to do it without any issues? Can you share a sample of the file that caused problems here (attach it to the thread), just to see if I can reproduced the error? I have used the name LONG for fields in the past and have had no problems. It is not a reserved word that cannot be used for field names in a FGDB: FAQ: What are the reserved words for Esri's file geodatabase?

Re: Bug with adding XY data?

shan_sarkar — Tue, 30 Jan 2018 12:35:53 GMT

Apart from the formatting in the field names there is a possibility that the Longitude might have been confused as Y and Latitude as X. That can cause the data to be compressed in a location.

Re: Bug with adding XY data?

HelenGoodchild — Tue, 30 Jan 2018 12:37:21 GMT

They all built their own spreadsheets, so had all chosen slightly different headers. I've tried to download their work and replicate it but I just can't, even if I retype the headers as they had them originally! They were all created in Google Sheets then downloaded to Excel format, so I'm hoping that wasn't the issue. If I get any further students with the issue I'll try and get an unsullied version of their work uploaded to check.

Cheers,

Re: Bug with adding XY data?

PanGIS — Tue, 30 Jan 2018 12:41:18 GMT

I agree,

it is true it can generate confusion but LONG is recognized automatically as X, from my "ADD XY" tool in 10.5.1 and always been in previous versions.

When I read that the point have been plotted elsewhere I always think that either the two field were inverted or a wrong coordinate system was chosen.

Re: Bug with adding XY data?

HelenGoodchild — Tue, 30 Jan 2018 12:57:25 GMT

Hi Pan, Yes that was my thought too, and one I always double check, but it definitely wasn't in these instances!

Re: Bug with adding XY data?

XanderBakker — Tue, 30 Jan 2018 13:01:20 GMT

And the correct coordinate system was assigned in the process and the coordinate system of the data frame wasn't different in such a way that a transformation is required?

Re: Bug with adding XY data?

HelenGoodchild — Tue, 30 Jan 2018 13:32:35 GMT

Yep - WGS84 and they were inputting Lat Longs. We had all the usual suspects in terms of plotting issues across the class in general (people using formulas, columns formatted as text, etc, etc, so I'm usually good at spotting the problem) I tried a few times myself in the class and had the same issues. It didn't work until I changed the names of the field headers in the spreadsheet and re-imported in exactly the same way. It was highly mysterious, and frustrating I can't replicate it!

Cheers,

Re: Bug with adding XY data?

RebeccaStrauch__GISP — Tue, 30 Jan 2018 15:46:17 GMT

Maybe they have some extra characters or spaces or some other issue within the field too. But to eliminate any issues with the software, try running one of the other students "good" spreadsheets on this this machine, and one of the "bad" spreadsheets on a machine that worked. But my guess is, beside the possible confusion on the names (although I use long/lat), maybe there are some additional characters at the end? or an extra dot, etc. For example, if using DMS, maybe there was a forth set of numbers that isn't showing (column not wide enough?).

Anyway, although I haven't tested my own on 10.5.1, seems like if it was a bug, it would have been reported already (popular tool). I would suggest uploading a sample of the bad spreadsheet so others here can test.

Re: Bug with adding XY data?

IanFrakes1 — Wed, 31 Jan 2018 22:15:37 GMT

I found if you use the words "XCOORD" and "YCOORD" in Excel coordinate field titles the software 'recognises' these as Longitude and Latitude coordinates respectively and they are automatically selected in dialog boxes.

Perhaps it's a coincidence and it's actually picking up the unique coordinate data type, we usually work in UTM coordinates. An ESRI developer might step in and solve this mystery.

Re: Bug with adding XY data?

Tolo__AdamTolo — Tue, 03 Jul 2018 08:30:09 GMT

I have also noticed this using ArcGIS Pro 2.1

Steps I took:

- Imported Excel spreadsheet into a table

- Display XY Data and selected X Field as Location_X_Coord and Y Field as Location_Y_Coord

- Gave it a layer name and select the projection GCS_WGS_1984

- Run

Points ended up in the wrong location.

After troubleshooting for a while I tried the same steps as above but with ArcMap 10.3.1 and it worked fine.

So I tried renaming the original fields I had to X & Y and it works fine now.

That's gotta be a bug?