Click here to see the SAS code.

I've been a volunteer on several archaeological digs, and I've 
seen the extremely detailed records they keep - recording the exact
x/y & depth location of every artifact.  So it was a natural step that
I combined my interest in archaeology with my interest in computer 
graphics and maps, and come up with a proof-of-concept that shows 
how a computer system could allow you to "view" the archaeology data
and 'reconstruct' the dig site.

First I create an excel spreadsheet containing a list of all 
the dig sites.  By listing all the info for the sites here, 
I'll later be able to just reference the site number later,
without repeating all the site info.

I use sas "proc import" to read the data from the excel spreadsheet 
into a sas data set.  You'll need "sas/access to pc file formats" to
read the spreadsheet directly into sas (and this only runs on a pc).
Alternatively, you could export the data from the spreadsheet into
an ascii file, and then base sas could read that ascii file.
I then merge in the county FIPS code numbers to go along with the 
textual conty names (this will allow me to match up my data to the
us county map).

Next, I created some "plausibly real" artifact data, and stored it in an excel
spreadsheet.  For each artifact, there is one row in the spreadsheet,
containing several columns of data (the columns will become my 'variables'
once they're read into a sas data set).  Each row contains enough info
to uniquely identify the artifact.  The spreadsheet contains artifacts
from many different dig sites.  I then use sas/sql queries to do some 
summary counts, such as the number of artifacts per dig hole, etc.

Now that I have the sites and the artifacts, I start creating some maps.
SAS already has the US map by state, and the state maps by county,
but once I get down to the "site" level I have to create my own custom
sas/graph map data sets.  So, from the main US map, click on a red-colored
state (such as SC) to drilldown to the state level, and then click on
a red county (such as Allendale) to get down to the county level.

At the county level, I "annotate" a red dot at the location of each 
"dig site" in that county.  I use annotate's "pie" function to draw
the red dot.  I use the annotate's "html" variable to code in my 
hmtl alt= charttip/flyover-text info, so that when you hover your 
mouse over the red marker you see the dig site's name (this works 
using the Internet Explorer web browser, but might not work in other
browsers such as netscape or mozilla).  I use the html href= tag to
code a drilldown, so you can click on a red dot and go to the map of
that site.  For example, click on the upper red dot to go to the 
"Big Pine Tree" site map.

For each dig site, I create a custom map showing the "holes" at that
dig site (I define a 'hole' as a 1-meter by 1-meter area, since that is
the size used at most of the dig sites I worked with.  Each hole is 
represented as a rectangle, by specifying the 4 coordinates of the 
corners (the coordinates are all in relation to an "origin" coordinate
for that dig site, as is recorded at each dig).   I use sas/graph's
"proc gmap" to draw the custom map, and I use gmap's "html=" option
to code in the alt= charttip/flyover-text (showing the site name, 
hole number, and artifact count), and the href= to do a drilldown
to show some graphical analyses of the artifacts in that hole.
Click on the middle hole, which has 22 artifacts, for example...

When you drilldown into a hole, I show several different graphical 
representations of what's in that hole -- a 2d top view, 2 2d side views,
and then a 3d view.  The 2d views are created using sas/graph "proc gplot",
and the markers are color-coded to indicate the type of artifact.
You can mouse over these colored markers to see that artifact's details
in the html alt= charttip/flyover-text info (if you're using the 
Internet Explorer web browser), and you can click on the marker to see
an image of that artifact (since it was difficult to come up with 
plausibly real pictures of each artifact, I re-used many pictures in
this proof-of-concept, but it could easily be set up to show a
picture of each individual artifact, if such pictures were available).
SAS/Graph "proc g3d" was used to show the 3d location of the artifacts
(g3d only has limited functionality, and does not support charttips
and drilldowns, and has some other limitations - it looks nice, but 
it does not have too many useful capabilities ;)

Below the graphs is a table listing all the artifact data in that hole.

---

There are *many* possibilities for expanding this proof-of-concept!

SAS offers many kinds of analytics, and they could easily be added.

Also, an interface could be added allowing you to dice-n-slice the
data ... for example, you could have it subset the data so that it
only shows the artifacts found between a certain range of depths,
or in a certain type of soil (A, B, C, etc).  Alternatively, it 
could be set up to allow you to pick a category of artifact
(such as clovis preforms) and show only those, and then maybe
show a plot of the whole site, so you can easily look for trends
in the location of certain categories of artifacts, to help plan
where to dig the next hole.  The possibilities are limitless!  :)