Where O Where Do I Open
a New Consignment Store in Vancouver?
Data Acquisition and
Manipulation
Store Locations:
I got the location of
existing
stores from the Yellowpages and the mytelus.com webpage. I
created an Excel sheet with the names and addresses and imported this
into ArcMap. Once in ArcMap I created an Address Locator and geocoded
the stores to the Vancouver shapefile off the SIS drive. This
took many tries to get above a 30% success rate. By the end I
only had to manually match 7 out of 48 locations. See Existing
Stores
I then had to group the stores together somehow to make their locations
meaningful. So I imported both the DotAreas shapefile (of
Vancouver neighbourhoods) and the geocoded addresses to IDRISI. I
then simply opened one map on top of the other and visually counted how
many stores were within each neighbourhood. For this part I
counted all stores for each neighbourhood, so if there were stores on
the edges these were counted. Proximity to people within the
neighbourhood was more important than assigning only one neighbourhood
to each store. See the map I used here.
I then used Edit/Assign to assign the # of stores to be the value of
each neighbourhood. This layer was then used in the final
multiple criteria evaluation.
Lease
Cost:
I looked at the Royal
LePage website at http://matrix.royallepage.com/ to find current
listings for commercial space for lease in Vancouver. I then
created an Excel sheet with the address and price per square foot for
all the listings. I then imported this into ArcMap and geocoded
it to the Vancouver shapefile with the same Address Locator I created
for the store locations. After this I imported the file to IDRISI
and used INTERPOL to create an interpolated image using Inverse
Distance Weighting. See the Cartographic
Model and the Interpolated Coverage.
I then wanted to mask out all areas that were not for
commecial use. So I used the Vancouver landuse file from the SIS
drive and selected for the Commercial and Commercial/Mixed Residential
areas in ArcMap. I then imported this file to IDRISI where I
created a Boolean image of Commercial and Other areas. See the Boolean Map
(with a vector layer for
Vancouver). I then overlayed the interpolated image with the
Boolean image to get a cost surface for the commercial areas only. See
the cartographic model (here) and the
resultant image (here).
Demographic: I acquired demographic data from
the
Census folder on the SIS drive. I used the census extractor and
extracted the age and sex data for 2001 into Excel. Once I imported
this into ArcMap I created fields to calculate the total male
population between the ages of 15 and 34 and the same for females for
each DA. I then created a field to calculate the ratio of women
to men and joined this table to the census shapefile on the SIS
drive. Once I had done this I imported the shapefile to IDRISI
and converted it to a raster image. From this I created two Boolean
images, one of DAs with a greater proportion of females and one of DAs
with a greater
proportion of males. To
make this data make sense for my question I opened each layer with the
neighbourhood layer and visually decided which neighbourhoods had a
greater proportion of women between 15 and 34. The DA's were too
small on their own, and I could not find a way to aggregate them into
the neighbourhoods utilizing the databases. See all three
maps here , both Boolean
images and the female image on top of the neighbourhood image. I
then used Edit/Assign and created a layer of only those neighbourhoods
that had a greater proportion of women.
NOTE - For all of the above
manipulations I first had to crop the images and databases so that I
only used information for Vancouver. This was done in ArcMap
using the 'Select by Location' feature in conjunction with the DotAreas
shapefile on
the SIS drive.