Select to view content in your preferred language

Question on how GIS works

4570
3
05-23-2015 02:45 PM
BarryPolinsky
Deactivated User

Greetings and Hello World!

I took a GIS course in my graduate MPA program and I found it to be very fulfilling, and I found that map making ability to be very useful. It adds a very technical element to what otherwise is a not so technical study, and it's a way that statistics, narratives, budgeting, policy, and other aspects of administrating in the public can work.

I have some questions about how GIS works however:

1) Upon booting up any GIS program, there is a user interface with buttons that trigger functions and output files in such ways. For instance, there are vector files, raster files, line files, and the like. If I import a vector file, how does GIS know where to place it on the screen? How does GIS know where to place it withiin the user interface, where it can be manipulated? Very often I'll import a vector file, or a raster file, or a line file, or any other file, and somehow whatever file that is either ends up focused or not focused on the screen in front of me and I have no idea why that is the case. Sometimes I'll import a file, it will end up no where that I can find it, and have no idea where the file is, so I just delete it and start over. There's obviously something going on in GIS which takes some element of the file that I am uploading or adding to GIS, puts it through some sort of process, and determines where to place it on the screen, Except: 1) I don't know what elements of the files GIS is looking for, 2) I don't know the underlying mechanics of how GIS assigns certain files or layers to take up certain space within the GIS context. Is this primarily the function of the Map Projection? For debugging purposes, how can I look at the various files that I am uploading and determine whether the issue is with the given files that I am uploading, the base file that I am using, the map projection, or something else?

2) I'm assuming that in some cases, GIS places things on the basis of longitude/lattitude and in other cases in places things on the basis of street address. For instance, it knows how to places a vector, raster, or a line file in a certain location because the file has certain longitude and lattitude built into it, and I'm assuming that GIS has the longitude and latitude system built into its underlying code. So if a given file has X longitude and Y latitude, GIS looks up X,Y in its internal long/lat coordinate system and matches the file accordingly.

     Furthermore, for the purpose of something like address location, I'm assuming that if someone has a base map of line files, and each of the line files has associated with it a name, a beginning number, and an ending number, and if someone tries to perform address location of a CSV or XLS file, GIS will: look up in the base file for the specific street name, fine the boundary conditions (for instance, if there it will determine that if address X is 85, and this given line segement has numbers that start at 80 and end at 90, a dot should be placed approximately half way from the start and end point). My question is: what is the process by GIS takes the information that I upload and puts it out in a given pattern? How can I determine what the process it is using is?

3) What are the differences between vector layers and raster layers? If I have a raster layer, how does GIS know how to place it with accordance to a vector layer, or not place it near? As I understand, a raster layer is a series of dots, or small vectors that represent a single point or series of points; vector layers can be geometrical in their layout. How do raster layers take up space at a given location?

0 Kudos
3 Replies
SepheFox
Deactivated User

Hi Barry, that is a lot of questions. For question 1, yes, it's a function of map projection. When you load the first layer into a map project, ArcMap will focus on that layer, and the projection of that layer will become your data frame projection. After that, layers are positioned in relation to that layer. If they have a different projection, they will be re-projected on the fly to the same projection as your data frame, but their original projection has not changed. If they have no projection, ArcMap will ask you to define the projection. If you do it wrong, they will appear way off the map somewhere. You can always look at the projection of your data layers, by looking at the the source tab in properties. If their is no projection assigned, you can make a guess based on the numbers listed for the layer extent.

For question 2, address data is generally geolocated, that is, those addresses are given coordinates based on the projection of the layer.

You will notice that shapefiles have a.prj file. That is here the projection info is stored.

Question 3, Raster layers are cell based, with each cell, like a pixel, having a values, and the layer being a grid of cells. Vector layers are made up of points with pairs of x/y location. Those points may be joined to gether to form lines or polygons. Have a read of this link. It might help: A Framework for GIS Modeling

SepheFox
Deactivated User

Hey Barry, I just noticed you have your question posted in ArcAPI. This is a place for java developers, and unlikely to get you the answers you need. I suggest moving it to New to GIS​. To move it, click on the title to open the question outside of your inbox, and select move from the right hand menu.

0 Kudos
hhet
by
New Contributor III

yes the process of rendering maps in a gis is a process of converting the map to numbers.  absolute location  might also be important to fixing the map to the analysis frame.  street addresses cannot be referenced without absolute location. you have already stated the standard ways to encode  earth locations [lat/long, utm, state plane, etc].

encoding maps requires .a conversion to numeric code.