problem when convert a x and y file to a point shapefile

3533
18
04-03-2015 02:26 PM
Highlighted
by
New Contributor III

I was trying to convert a group of x and y coordinates into a point shapefile. I use ArcMap 10.2, Here is what I did:

(1) put x and y (in numerical values) in columns in Excel and convert it into a .cvs file;

(3) for coordinate system, I imported the projection of my reference shapefile;

(4) click ok.

(5) export the layer into a shapefile by right click it and go to data>export data;

After I got the shapefile, I realized that it cannot be in line my reference map though both have the same projection. When I check the converted shapefile, I noticed that its EXTENT is different from that of the reference map.

Could anyone tell me what I did wrong and how to fix this problem?

Many thanks.

Peng

1 Solution

Accepted Solutions
Highlighted
by
Frequent Contributor

Well, that's true for lat/longs Dan, but what about UTMs. Ah, isn't it fun being a coordinate system geek!

18 Replies
Highlighted
by
Frequent Contributor

It is important that the projection you assign to the x/y points is actually the projection of the coordinates, not the reference file. Do you know what kind of coordinates they are? Are they Lat/Long coordinates? UTMs? Other? Another common mistake, at least here in North America is to forget that coordinates can be negative.

Highlighted
MVP Esteemed Contributor

In addition to the other comments... people often switch the longitude and latitude values around because of the propensity to say latitude-longitude, thereby screwing up the proper reference to the X and Y axes.  And to clarify, only the longitude values in North America will be negative.

Highlighted
by
Frequent Contributor

Well, that's true for lat/longs Dan, but what about UTMs. Ah, isn't it fun being a coordinate system geek!

Highlighted
MVP Esteemed Contributor

actually eastings and northings is the correct way to reference (ie X and Y) and there are no negative values, (the southern hemisphere denotes the equator as 10,000,000m and works down.  however, there are many projections where the X value can be -ve depending upon the central meridian used...ie Canada's use of the Lambert Conformal Conic projection which is widely employed by Canadian Governments.  the sad thing is that most people don't realize that when you get values in the range +/- 180 and +/-90 you are probably working with a geographic coordinate system and not a projected one

Highlighted
by
New Contributor III

Thanks for you both. I think I got lat/long correct. After conversion, I can see that the converted point shapefile ahs the same projection as the one in the reference map, However, when I go to 'properties' of the converted point shapefile, I noticed that its extent is still based on the original values of lat/long, though there is unit 'ft' at the end of the value. I feel that I may have to do something to make sure that the extent will also be changed to that of the projection I used, but I do not know what to do?

Any suggestion? Many thanks.

Highlighted
by
Frequent Contributor

Peng, it still sounds like you made the projection match your reference map, instead of whatever they were originally. Is that right?

Highlighted
by
Frequent Contributor

Can you send a screenshot of what you are seeing in the properties?

Highlighted
by
New Contributor III

I think you might be right. I have to go back to check how the lat/long data were obtained. Following is the screenshot. Note at the righ bottom, the unit is still degree though I have already defined the projection for this data.

?[cid:db6044ee-aebb-475f-b084-da05278875ff]

Highlighted
MVP Esteemed Contributor

your extent definitely suggests longitude/latitude values and the definition to state plane is wrong.  Did you use the Project tool, in the Data Management tools, Projections and Transformations...or did you use the Define Projection tool?  If you did the latter, you defined it wrong.  When you physically project data you will end up with a new file, if you define it, you just put a suit on a dog...but it is still a dog and nothing has changed it.  the procedure is

1. define the coordinate system of the data...it is either a geographic coordinate system (aka long/lat) or a projected one...that you would have to know...no guessing...find out
2. once defined, then you can Project the data to get it into the coordinate system you want
3. this should all be done in a dataframe with no other data in it, because the first file that you add, sets the coordinate system of the data frame and when you project it...nothing will look like it has changed because the 'evil...projection on the fly' takes over and tries to help you.
4. once projected, open a new dataframe and add the freshly projected data...it will look different and the dog will no longer have a suit on