prob

* My first idea was to analyse the distribution of 'Starbucks' in Vancouver. The next step was more difficult: What did I want to analyse in detail? I thought about the relation between 'Starbucks', Malls and Skytrainstations. But after I had decided to do so, I was told that this would be too simple. The next idea was to compare all this in relation to the distribution of 'Blenz'. Therefore, I digitized the location of all the 'Blenz' in Vancouver. But having the final idea comparing different social factors with the distribution of 'Starbucks' I couldn't do the same analysis for Blenz in time.

* When I digitized the location of the 'Starbucks' for the first time, they were too small. As a result they weren't really visible on the map. To avoid this effect I chose a bigger size. Now, you can see them on the screen, but the precision is not that good.

* The data I got had information about how many persons are living in one household. The problem was that one of the categories showed both, how many 4 and 5 persons are living in the Census Tract. So, I couldn't make a difference between them.

2. Operational Problems:

* The first data I collected had information about the Enumeration Areas of Vancouver. I thought that smaller units would give a better overview of the area. After manipulating them in ArcCatalog, Excel and joining them in ArcView 3.2, the last step was to project the map into NAD_1983_UTM_Zone_10N. And there the problem occurred: The units were to small and after the PC calculated several times for about half an hour, it always crashed. I was told that the units are too small (the information of the data was too much) for a projection. At this time I had to start with the data collection again and it needed a lot of time to get new ones. At least I got the data of the Census Tracts. The precision is not as good as before, but the projection worked.

* After I finished the maps with ArcView 3.2 I had to transform them into IDRISI. The problem was that the joined attribute table with all the data of the Census Tracts was not transformed. I tried different operations to get the lost information before I knew that I had to save the joined table as a shapefile in ArcView 3.2.

* The next problem I had was probably my fault: I had a rewritable CD, where I stored all my data. But it seemed to be of a bad quality and one day I couldn't open the data, because several PCs didn't find the CD itself. My luck was that some of the data was also stored on my h: drive. But I had to repeat a lot of work, which I stored on a new CD...

* I don't know if it was only me, who had the next problem or if the function does not work with vector-formats, but when I overlaid the layer of the different Census Tracts and the layer of the distribution of the 'Starbucks', I couldn't save them as a "Map Composition". The PC always told me 'vector file not found'. This is the reason why my basic map has a raster surface. I preferred the vector objects, but it was not possible...

* I created a new basic map in ArcView 3.2, because I don't like the layout of the raster-map. I stored the map first as a *jpg-file and then as a *gif-file. But the quality of both of them was very bad and I didn't want to take them for my presentation.

* The next function I don't like about IDRISI is that the converted raster-format map of the Census Tracts couldn't be connected with the attribute table.

* At least I didn't enjoy working with Netscape Navigator: First, it provides too few options creating the layout. Second, it always do things it shouldn't making no sense: Usually it doubled the letters while writing cursive or in bold. Deleting only the letters means to delete all. Sometimes it changes the space between the lines and you can't do anything ... and sometimes it crashed...
Because of the latter I had to some of the work more often, which wasted a lot of time. At the end I learnt my lecture and stored after every single sentence...

5. Methodological and Operational Problems