Preparing our data

From the currently opened vector layers, the GeoNames layer has the largest attribute table with the most kinds of attributes. However, as the extract does not contain headers, it is quite hard to work with it. Fortunately, CSV files can be edited as regular text files or as spreadsheets. As the first step, let's open the GeoNames file with a text editor and prepend a header line to it. It is tab delimited; therefore, we need to separate the field names with tabs. The field names can be read out from the readme file in order. In the end, we should have a first line looking something like this:

    geonameid name asciiname alternatenames latitude longitude
      featureclass
    featurecode countrycode cc2 admin1 admin2 admin3 admin4 population
      elevation
    dem timezone modification
If you use a text or code editor which replaces tabs with spaces, don't forget to switch off that feature before adding the header line. Furthermore, copy-pasting the preceding code block will probably not work.

Now we can remove our GeoNames layer from the layer tree and add it again. In the form, we have to check the option First record has field names. If we do so, and name the latitude and longitude fields accordingly, we can see QGIS automatically filling the X and Y fields:

You can speed up rendering and spatial querying by checking  Use spatial index. It will consume some memory though, so be careful with unusually large files. If you have such a big GeoNames layer that you cannot work with (like the whole U.S. table on a weaker computer), you can choose another layer for the next part.
上一章目录下一章