As with any GIS project a large portion of time is consumed by the processes of data collection and decision on what to include in your analysis in order to display the appropriate information. For this GIS project, data was collected from a variety of different sources. . Through the census analyzer data such as total population density of 2001, average income, unemployment rate, and immigration population were chosen from the Dissemination Area (DA) for the BC – Greater Vancouver Regional District (GVRD). After the data was collected from the census analyzer the data converted as a database file (.dbf) where a primary key such as DAUID was created in order to be compatible for use to join database files to map shapefiles in ArcMap. Majority of the secondary data for this project was collected from various folders, such as DMTI Datasets, GVRD, GVRD Landuse, along with other data were from the data warehouse in the S:/ Drive of the SIS Network at Simon Fraser University Burnaby Campus. In order to receive permission to access the DMTI Datasets in the SIS Network, an agreement form had to be filled out for security reasons. There were secondary data that was collected for analysis that had to be collected from outside sources because it was unavailable in the data warehouse in the SIS Network. For example, a map shapefile of Surrey that displayed city district boundaries was not available in the data warehouse. In order to obtain this shapefile, it was necessary to contact the GIS Products and Services section of the Engineer Department for the city of Surrey to request for this dataset. A form was filled out and sent to them. After contacting the Surrey GIS Products and Departments: 604-591-8693 and working out legal issues with them, the GIS department sent the following vector shapefile: Data Preparation is also very important. First, different map shapefiles and layers had to be displayed and prepared in vector format using ESRI ArcMap. Then the maps were joined with necessary database files in order to output census data on maps for such topics as average income or unemployment rates. In order to confine the maps to show just the area of interest for this project, Surrey was clipped out of the GVRD map. Some of the primary data entry was also done at this stage, such as digitizing the existing police station in Surrey. All Vector maps used the GCS North American 1983 Geographic Coordinate System and NAD 1983 UTM 10N projected coordination system. One exception was for the Surrey Communities where the coordinate system for the shapefile was undefined upon receiving the shapefiles from the GIS section of Surrey, therefore the NAD 1983 UTM 10N was given as the projected coordinate system for the Surrey Communities shapefile as well. One of the most important steps for data preparation in this project was to ensure that all vector shapefiles were clipped according to one map with the largest area, in this case it was the landuse map of Surrey, and then the other map layer would then be set to the same area and grid size. This is important because this will enable the grids to match up when attempting to run analysis for raster format files inside IDRISI. A grid cell size of 70 was given to each of the different maps used for this project. The raster clip function in ArcMap was used to clip all other shapefiles to the same area of extent was the landuse map. This is shown below: When all the shapefiles were clipped to have the same extent as every other shapefile, the next step was to convert all the vector shapefile into raster file format, so analysis can be done in a raster based software like IDRISI. The following image shows an exaggerated animation of the process of converting from vector file format to raster file format: Another important part of data preparation for this project was reclassing the raster maps created in ArcMap inside IDRISI. The raster maps needed to go through reclassing in order for raster map layers to make sense and correspond with one another for analysis inside IDRISI. One of the reasons for reclassing each raster map was to eliminate the negative (-9999) values. It is also necessary to reclass files so that certain names in the attribute table match the code numbers of the attribute tables. For example, in the landuse map it was reclassed so the landuse code that was given to each landuse type in the attribute table matches the landuse type names in the attribute table. For instance, the landuse code 1 in the attribute table is for land used for transportation, communication and utilities. Below is an example of the landuse map after it has been reclassed in order to eliminate the negative (-9999) so the landuse code correspond with the landuse name. Refer to cartographic model link above to view the exact numbers used for reclassing the landuse map. Source.