Time for my annual post! I’ve decided that instead of chatting generally about my Ph.D., I can use this site to record tutorials on the geographic and statistical techniques I am using. Hopefully someone finds this interesting, but it is also a useful public record of certain techniques for future reference for me or students. Today I worked out (quite simple, really!) how to make a simple choropleth map in QGIS 2.14 (Essen):
Step 1: Start with two vector data types: one of points, the other of polygons. I have the metal-detector finds from all periods for Sussex (from the PAS database) and the digitised 1851 parishes from Kain and Oliver (200?). Note that some of the northings and eastings are incorrect!
Step 2: In the menu, choose Vector > Analysis tools > Points in Polygon
Step 3: Choose the layers you want to count. Your polygons are going to be the containers in which your points will be assigned. You can customise the name of the field which the sum of the points in the polygon will be recorded. You can also give your .shp file a name. When you’re ready, press ‘OK’.
Step 4: Voila! …Nothing happened. You get a new .shp file that looks exactly like your old one. If you open the attribute table, however, you’ll see the new field and numbers which represent the counts of points in each polygon.
Step 5: While we’re in here, we will replace all of the ‘NULL’ values in the field to ‘0’. We can do that with the ‘Select features by using an expression’ button. In this case we want to select all of the rows where the field “PNTCNT” is NULL. Press ‘Select’ and then ‘Close’. Go back to the attributes table.
Some rows should be selected. Make sure that the layer is editable (the pencil icon in the upper left corner) and that the field you want to edit is selected in the drop-down selection box. Put ‘0’ (without the inverted commas) in the box and click ‘Update Selected’.
Step 6: We can close out of the attribute table and go into the properties of the layer we created. Right click the layer in the ‘Layers’ panel and select ‘Properties’. In the style menu, we select ‘Graduated’ from the symbol selector at the top.
Step 7: I chose the blue colour ramp because it’s my favourite colour (inverted means the darker colour will be closer to 0, but as density of points it makes sense that the higher the quantity = the darker the colour. Feel free to pick any of the other colours or invert as you wish! Check out http://colorbrewer2.org/ which is a web tool for exploring the use of colour in map-making.
In this case, the map is less than helpful, as there has been an awful lot of metal detecting in only a few parishes which drown out the results. In my case, the inverted option makes more visual sense. Experiment with other classifications to see which best represents your data. In this case a heat map would probably be a better option than choropleth, but it is best to consider what combinations of raster and vector data is more suited to the type of information you are trying to communicate.