QGIS, I could kill you.

I’ve been dreading taking on QGIS since it was assigned in my Digital Methods class at the end of September. I ran through the tutorial then, but was perplexed by the lack of intuity for the application. I just didn’t understand it beyond following the step by step instructions. I decided that I would put it off until I had more time to dedicate to it. In hindsight, this seems ridiculous. Since I have returned to school there has not been a time when I had more free time as I progress through a semester. I did have reason to believe this semester would be different though. And, the ultimate irony is that I now understand how powerful this software is and I even kinda like it.


It took two days, and here’s how it happened. I went back to the beginning. I started with a basic shape file from the City of Albuquerque public data set. I used city council districts so that even with this first layer of mapping I would have some shape over which to apply more. And I did. I added city roads, then parks. As the layers intersected, I could see that I should be able to build a map that is more pertinent to my interests.


I started doing web searches for Indian lands, hoping to find some data from the 19th century in the Southwest. I didn’t find that. But, I did find a USGS website of small-scale data for the national map. There is lots of cool stuff here, including a shape file of Indian lands from 2005. OK, so what could I do with this? I laid down a base layer of U.S. states, then the Indian lands and then waterways, roads, and wilderness preserves. Nothing was striking. What I was missing was points on my map that would tell some story. I’m also interested in public health, so I decided to look up the addresses of New Mexico’s Indian Health Service clinics and see where they would lie on the map.

But how to get address into coordinates? I started searching for Python coding that may do this, and instead found a plugin for QGIS called MMQGIS that has geocoding capabilities. It took a while for me to get it installed, but once I did I could convert a csv file to a layer of points on my map. MMQGIS has some cool functionalities that I did not get into, but I will revisit in the future, like animating points to show movement. Anyhow, I used the IHS addresses to create a csv file in Text Wrangler and ran it through MMQGIS. I had to do it three times. The first time I got several returned lies because of misplaced commas. Commas are important for csv. Then, I kept getting one address returned as not geocodable. It is on the Santa Ana Pueblo and a bit irregular.

What I learned about even such a simple project, is that IHS hospitals and clinics are overwhelmingly placed on Indian land in New Mexico. The few outliers are in the large urban areas. Navajo Nation clinics are not mapped here.

This is my fail for the lesson. I searched for a good address on Google Maps and found the facility but couldn’t get an address. I rewrote the address from IHS’s website different ways. I just couldn’t get it to work. So, with nine out of ten IHS clinics in New Mexico mapped against Indian lands in the state, I gave up. It is an overall success because I now feel like this software is not beyond me, but I’m still a little disappointed. I’ve thought about how to overcome this problem in the future. I could use Google Maps to get coordinates of the address that won’t compute and enter it in separately. I think there must be an even better way too, but I was beat. It was 1 a.m. Next time I use QGIS I will allow time for contingencies like this. I wish my map had ten points.

Show your support

Clapping shows how much you appreciated Kaveh Mowahed’s story.