methodology
*After viewing images, simply hit back on the browser to go to the particular section you were on in this page or else it will reset to the top which can be annoying.
(1) To begin my analysis, the first constraint was that the new snowboard shop had to be in a commercial area. With the Landuse in the GVRD raster image already available in IDRISI from the network (See Image), I simply assigned all commercial areas as 1, and non-commercial areas as 0. See Image.
Then, in IDRISI using buffer, I set a 500 m buffer zone around commercial areas so that suitable locations for a new snowboard shop can be located to a maximum of 500 m from the commercial area (a walkable distance). See Image.
(2) Next, I had to create a layer showing the population of males and females aged 15 to 30. I also decided to show only those areas over 200 people per enumeration area meeting this criteria. I figured this would be more accurate if the shop is placed in a more populated area to be a successful business. In order to create the first layer, I used an ArcView shape file on enumeration areas of the lower mainland from the network. I, then, used census data from the Research Data Library webpage to link this information to the shape file. This census data was in a .csv file and was brought into Microsoft Excel for editing. I was able to delete all unnecessary columns and rename them to be identified in ArcView. I saved it as a DBaseIV (.dbf) file to be used in ArcView. In ArcView, I opened the attribute table for EAs shapefile and selected the EaNumber field. I, then, joined the datasets together saving the file as a newea.shp file. My next step was to convert this into IDRISI in a vector layer then as a raster layer. It was here that I ran into my first problem. After converting it into a vector layer using FME, in IDRISI, the layer wouldn't convert to raster using polyras with my commercial raster layer as a base. It said that the datasets did not match. A friend of mine suggested the only way he could think of making the newea.shp into an IDRISI raster layer was to use ArcInfo. He was able to help me convert it into raster. See Image.
After this, I created my final population layer showing areas with over 200 people of males and females aged 15 to 30 by using reclass. See Image.
(3) Next, I had to create a layer showing average household income of at least $50000/year. In order to do this I did exactly the same procedure as in #2 for my first layer except I opened the new attribute table for newea.shp file and joined the datasets together saving it as the same shape file. I was then able to convert this into raster as a new layer with the help of ArcInfo. See Image.
After this, I created my final income layer showing average household income of at least $50000/year by using reclass. See Image.
(4) Next, I created a layer showing the streets in the Vancouver area by using a data set that was available from the network as an IDRISI vector layer which had to be converted into raster. In IDRISI, I used lineras to convert it. I, then, had 3 values where one was not roads and the other two being roads. I assigned both roads as 1 for being roads and 0 for not being roads in order to do my final analysis. See Image.
(5) Next, I created a layer showing the Skytrain line. This layer was available from the network as an ArcView shape file which had to be converted into IDRISI. I used a translator software program called FME to do this. I then had to convert the vector layer into raster for my final analysis. In IDRISI, I used lineras to convert it. See Image.
Then, in IDRISI using buffer, I set a 500 m buffer zone around the Skytrain so that suitable locations for a new snowboard shop can be located to a maximum of 500 m from the Skytrain (within walking distance). See Image.
(6) Next, I had to create a layer showing the existing locations of snowboard shops in the area. Data showing competing shops was not available, so I manually digitized the points as a vector layer on the road raster layer. To do this I was able to consult the BC Yellowpages on-line to verify locations of existing shops. In IDRISI, I digitized each point based on the approximate location. This process was not entirely accurate as the street names were not available on-screen and the resolution being poor. The shops I considered to be competition for the purposes of this project included The Boardroom Snowboard Shop, Pacific Boarder, Westbeach Snowboard Canada Ltd., Comor-Go Play Outside, Pain Snowboards, Warp Board Culture, Ride on Sports, Thriller, The Boarding House, and Skyline Sports. The result after digitzing was that a point vector layer was created. I, then, used pointras to convert this layer into its own raster layer. See Image.
To create my final existing shops layer for suitability, in IDRISI using buffer, I set a 1000 m buffer zone outside the existing snowboard shops so that suitable locations for a new snowboard shop wouldn't be located too close to competition (within 1000 m), but within walking distance of one. See Image.
main | conceptual outline | data collection | methodology
spatial analysis | problems & errors | about the author