Skip to content

Commit 99e3d77

Browse files
committed
Extend manual section about discrete isolation generalization
1 parent aa2589a commit 99e3d77

File tree

1 file changed

+19
-6
lines changed

1 file changed

+19
-6
lines changed

_manual-v2/10-generalization.md

Lines changed: 19 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -160,7 +160,7 @@ parameters):
160160
| Parameter | Type | Description |
161161
| ----------------- | ---- | ----------- |
162162
| id_column | text | The name of the id column in the source table. |
163-
| importance_column | text | The column in the source table with the importance metric. Column type must be a number type. |
163+
| importance_column | text | The column in the source table with the importance metric. Column type must be a number type. Numbers must be positive, larger number means more important. |
164164
{:.desc}
165165

166166
The `src_table` and `dest_table` have always to be the same.
@@ -173,9 +173,9 @@ configurable:
173173

174174
| Column | Type | Description |
175175
| --------- | ---- | ----------- |
176-
| discr_iso | real | Discrete isolation value |
177-
| irank | int | Importance rank |
178-
| dirank | int | Discrete isolation rank |
176+
| discr_iso | real | Discrete isolation value. This is the smallest distance from this feature to all other features on the map in map units. |
177+
| irank | int | Absolute importance rank derived from `importance_column`. Most important feature gets 0, next important one gets 1, and so on. |
178+
| dirank | int | Discrete isolation rank derived from the discrete isolation value. Most important feature gets 0, next important one gets 1, and so on. |
179179
{:.desc}
180180

181181
Use these column definitions in your config file to add them:
@@ -186,9 +186,22 @@ Use these column definitions in your config file to add them:
186186
{ column = 'dirank', type = 'int', create_only = true },
187187
```
188188

189-
See [this blog
189+
When deciding what features to render on some map, you can take the following
190+
into account:
191+
192+
* The `irank` value, the absolute importance of a feature.
193+
* The `dirank` value, the relative importance of a feature.
194+
* The zoom level or scale of the map. Basically this decides how much space
195+
there is on the map to put stuff into.
196+
197+
You can give different weights to the absolute and the relative importance,
198+
respectively. See [this blog
190199
post](https://blog.jochentopf.com/2022-12-19-selecting-settlements-to-display.html){:.extlink}
191-
for some background.
200+
for some background and an example formula.
201+
202+
This strategy is based on this paper: [Scale-Dependent Point Selection Methods
203+
for Web Maps](https://dx.doi.org/10.1007/s42489-021-00079-y){:.extlink} by
204+
Mathias Gröbe and Dirk Burghardt.
192205

193206
#### Strategy `raster-union`
194207

0 commit comments

Comments
 (0)