I’ve completed my project write-up and presentation, so I figured I’d release it to the wild. Chances are pretty good no one in my class will see it before I present and spoil it. Actually, I’m more worried that this may reduce our chance of actually finding one of these places.
The gist of the project was trying to find areas in Portland that meet the ideal requirements we have for a home using GIS. We compiled a list of things we wanted and things we didn’t want and then looked to see if there were actually any affordable houses in those regions. Here it is:
- the GIS I Project Summary
- the slide show
Both are in PDF format. I’ll have a pretty final map to include tomorrow.
Andy, this looks great! This would have been useful when we were looking for a house in PDX three years ago.
If you want crime data, check out portlandmaps.com. I have no idea how it would factor into the fancy mapwork that you do, but if you enter an address, the website will show you a map of various crimes in the surrounding area.
As for assessed value v. market value, that one is a puzzle even to the taxman. Assessed value and the true market value (as opposed to the “real market value” shown on tax rolls) differ greatly. For example, the house we owned on 81st had an assessed value of $87,970 in 2003. We sold it in 2004 for $146,500. The house two doors down from us had an assessed value in 2003 of $80,230, and it sold in 2004 for $132,000. The house three doors down from us had an assessed value in 2003 of $73,040, and it appears to have sold in 2005 for $134,000. Granted, these are just three properties, but it shows you how assessed value is not a very accurate measure of true market value. Even the “real market value” shown on the tax rolls is a bit off. The three properties I described had “real market values” of $132,400, $120,000, $112,760, respectively.
None of that, of course, takes away from the brilliance of your project. RMLS and the like are only useful to a certain point. They do not allow you to search by age of the home or by personalized neighborhood amenities (i.e. pubs, absence of major streets). Something like what you have designed would save a lot of driving past houses that a person would never want to buy, but can’t tell enough about to know that.
In any event, good luck with the presentation!
Cheers,
Derek
Fascinating. How fun to see your cool work. Susi
Looks really cool man!
This looks awesome. I’m glad you have a website with your projects. Great to see since I am just starting out in GIS. P.S. Do you happen to still have the shapefiles for your Where to live in Portland project?
I’m trying to figure out a way to share GIS data for things like brewpubs and will post back when I can find an efficient way to do so. At this point the data from my project is all in geodatabase format and is WAY to big to e-mail. I’m really starting to like this geodatabase format. Everything is so tidy.
Congratulations, very interesting project. I work for the City of Portland Corporate GIS, sorry we didn’t get back to you on your Crime Data request. Best place to go for that is direct to the Police Bureau. They may want to charge you for an export, but I think they also have public datasets available at the Neighborhood level.
The data in CrimeMapper can’t be directly exported because it contains sensitive information that could be linked back to victims.
Anyway, good work!
Later,
-Phillip
p.s. for access to the City’s GIS go to: http://www.portlandmaps.com
Thanks for the comments Phillip,
I figured that the crime data must be kind of sensitive and somewhat obfuscated in some fashion. What I was thinking of doing, sort of a poor man’s hack, was to create a generic quantified polygon shape made from eyeballing the overall crime map and using it. I’m mostly interested in the overall range, so specifics aren’t really that important.
That looks like a great project – nice work.
I am just a GIS user who happened upon your site and looked over your work.
In lieu of an “erase” feature, you can also get the same results (if I understand correctly what you wanted to do) by performing a Union. This would give you 3 feature sets: one of areas within layer 1, another of areas within layer 2, and a third within both layers – think of the mastercard symbol. Then you can keep or throw out any combination of those 3 subsets.
Have you tried doing any of this in a raster-based software like ESRI’s Spatial Analyst, or Clark Lab’s IDRISI? They have lots of cool ways to aggregate seperate contributing factors with relative weights, etc.
Good work!
Jeff