What is GIS?

So in a more formal sense esri's definition of GIS is, 

An integrated collection of computer software and data used to view and 
manage information about geographic places, analyze spatial relationships, and 
model spatial processes. 
A GIS provides framework for gathering and organizing spatial data and 
related information so that it can be displayed and analyzed. 
I think that second sentence is less dense and also a really good summary on its
own. 
And esri is the maker of ArcGIS, so they know a fair amount about GIS.

Still somewhat formal but a little more condensed, 

a different definition of GIS from Burrow 1986 is, 
A set of tools for collecting, storing, retrieving at will, transforming and 
displaying spacial data from the real world for a particular set of purposes. 
These are still really broad if you're noticing that trend. 
It could be a whole lot of things. 
A more practical definition of GIS would be, Software and 
data that enable us to ask and answer questions involving where something is and 
how that location relates to other things. 
So I'd say if it can analyze spatial relationships or 
if it can display location information to you, it's probably a GIS. 
That includes Yelp, Google Maps, Google Earth, your GPS in your car, 
anything that involves spacial information is probably a GIS.

Components of GIS, potentially, are design,  work flow and organization, how your
data moves through it and how you analyze it,  the specifications of your data or of
your software, the technology involved. 
So that would be from the computing hardware to the software to 
what the user interacts with. 
The algorithms involved, which ties back to the workflow and organization. 
And then the research that goes into it. 
This stuff is built on years and years of research on how computers work on 
how location information is gathered and processed and used. 
So, there are all sorts of things you should be thinking about in a GIS. 
And that's where you come in is without humans to integrate all these pieces, 
there's no system. 
GIS is geographic information systems, and the system is whatever we make of it. 
So when you put these components together, you make a GIS.

So what types of questions can we ask of a GIS? 

And when I talk about asking questions of a GIS, I mean that you have data that you 
want to learn something about and so you can present it to your GIS and 
put it through an algorithm or a work flow in order to get a new result out, 
and that's your answer to your question. 
So what exists at a certain location would be the most basic question you could 
ask in a GIS. 
Just, what's there? 
Where are certain conditions satisfied? 
What has changed in a place over time? 
What spatial patterns exist? 
What if this condition occurred at this place? 
And that's where you get into future prediction and hypothesis testing. 
And then, where do variables interact? 
There are tons more questions you can ask but notice that location and where and 
spatial is a really common pattern here. 
That's because it's a GIS, right?

So points build lines, lines build polygons, with points being connected by an edge
to make a line and lines being connected all the way around back to the origin to
make a polygon. Points are dimensionless. They are represented with size, but they're
infinitely small in practice.

Lines are one dimensional. They move in either direction. And each segment of a line
can have an attribute record, just like each point can have an attribute record. And
then polygons are by default two dimensional and each polygon can have an
attribute record. One thing to note though is we can add more dimensions to each
of these. We can add a z attribute to them so that they have height data. And then
we can also add time data so that. This attribute or the polygons with the lines or the
points have more or different information through time and so you can time enable
your layers. Raster data is the counterpart to vector data. We talk about GIS data as
being either vector data or raster data. There are a few other types but principally
these are the two that you're going to use. Raster data you've probably seen before,
it's the type of data that your digital camera takes. It's best used for imagery like
digital images or continuous surfaces, and what I mean by continuous surfaces is
where vector data describes ocean, sea, land like I talked about. Raster data can be
those numbers filling one to a million and any decimal place in between. So each
location can be really related to the ones next to it because they vary continuously.

Raster data is composed of a regular grid of pixels with fixed cell size. So let's deconstruct that a little
bit. If you want to think of a regular grid, just draw a Tic-Tac-Toe box for yourself and if all the
squares are of equal size you have a raster there. And if you write different numbers in them that's
how raster get their values and if I write a five in the top left box of the tic-tac-toe box. That whole
area that that raster cell represents has the number five as its value, whatever that number means
in this case. If it's elevation, maybe it's five meters tall. And then the one next to it might have a six
and so that ones six meters tall and the one below it might be eight, something like that. And so
Rasters do that through the whole space that the Raster covers.
Imagery is also raster data, just like we talked about. But it's oftentimes multiband raster data where
you have multiple values per cell which represent the colors of the color spectrum. So red, green,
and blue each have their own values per cell.

Geodatabases are where we hold our spatial data. You can hold it in other ways, but geodatabases
are nice containers to bring all of your spatial data together. Geodatabases are a general concept,
there are multiple types of geodatabases that you can have. You may hear about file geodatabases,
personal geodatabases, or spatial light geodatabases ArcGIS can use all of this and you can use them
in other systems too sometimes. Geodatabase helps us bring together both feature classes and
rasters. Feature class is being vector data. But also our tabular information, our relationship
information and our annotation information and even more.

Shapefiles which you may have heard of if you've done a little bit of work with GIS systems before
Shapefiles are often talked about as if they mean vector data in general but they don't. Shapefile is a
specific ESRI format. It's really common, it's a really important format because it's old enough and
widely published enough that it's used as an interchangeable format between different GIS systems.
So Shapefiles can be as few as three different files to make up a Shapefile and as many as seven files
on the disk. So if you happen to send them, need to make sure to grab all of those. So they're a little
unwieldy. The other thing is that in being old, they rely on old standards like dBase IV for their data
tables, and that makes them pretty limiting in terms of what you can do in your attribute tables. I
find them to be slow and large compared to other data formats, but they're incredibly useful and
incredibly important because that's how you often can send data to others you're working with. If
the other people you're working with are using ArcGIS and also have the knowledge to use the
geodatabase you can send it to them that way. But Shapefiles are pretty important for interchange
otherwise.

So to sum up all of these terminology, we have vector data and we have raster data, those are two
primary data formats. Vector data is best suited for categorical or discrete data. And rasters are best
suited to continuous data. Attribute tables are and especially important for vector data and to attach
additional information, but we can also attach them to raster values. And then geodatabases
commonly store our data on disc. And then map documents and layers allow us to represent our
data visually for cartographic purposes or to provide a workspace for analysis of our data.

 Spatial Data is a data with a geographic location, representation, or reference point

that it describes

 Components of GIS, potentially, are design, work flow and organization, how your
data moves through it and how you analyze it, the specifications of your data or of
your software, the technology involved. 

 Points are dimensionless. They are represented with size, but they're infinitely small
in practice.