New indicators to investigate LCZ classification #937
Conversation
… a territory: spatial distribution of objects, measurement of heterogeneity, neighborhood analyses, distribution of objects according to geographical directions, distances between objects...
| tablesToDrop << lcz_count_lod | ||
| datasource.execute(""" | ||
| drop table if exists $lcz_count_lod; |
There was a problem hiding this comment.
indentation ? For readability reason, would be better further right
| tablesToDrop << lcz_count_lod_mode | ||
| datasource.execute(""" | ||
| create index on $lcz_count_lod(ID_ROW_LOD_${i}, ID_COL_LOD_${i}); |
| tablesToDrop << grid_lod_level_final | ||
| datasource.execute(""" | ||
| CREATE INDEX on $lcz_count_lod_mode(ID_ROW_LOD_${i}, ID_COL_LOD_${i}); |
| * Create a multi-scale grid and aggregate the LCZ_PRIMARY indicators for each level of the grid. | ||
| * For each level, the adjacent cells are preserved as well as the number of urban and natural cells. | ||
| * To distinguish between LCZ cells with the same count per level, a weight is used corresponding |
There was a problem hiding this comment.
a weight is used TO SELECT ONLY ONE LCZ TYPE, corresponding...
| * @param datasource connection to the database | ||
| * @param grid_indicators a grid that contains for each cell the LCZ_PRIMARY | ||
| * @param nb_levels number of aggregate levels. Default is 1 |
There was a problem hiding this comment.
Is the default value 2 instead of 1 ?
|
|
||
|
|
||
| /** | ||
| * @author Erwan Bocher (CNRS) |
There was a problem hiding this comment.
suppose to be in spatial unit class ?
There was a problem hiding this comment.
This class will be deleted
There was a problem hiding this comment.
Why a new class ? I suppose having those in grid indicators would make the job ?
There was a problem hiding this comment.
If I understand well, everything will be moved in the spatial unit class ?
There was a problem hiding this comment.
It was the initial code. All methods will be moved in the good place.
| void multiscaleLCZGridTest() { | ||
| //Data for test | ||
| h2GIS.execute(""" | ||
| --Grid values |
There was a problem hiding this comment.
Not sure but is this test have an equal number of LCZ for one situation and thus has to use the weights to check that the weigths are OK ?
| @@ -225,4 +227,128 @@ class SpatialUnitsTests { | |||
| def countRows = postGIS.firstRow "select count(*) as numberOfRows from $outputTable" | |||
| assert 100 == countRows.numberOfRows | |||
| } | |||
|
|
|||
| @Test | |||
| void sprawlAreasTest1() { | |||
There was a problem hiding this comment.
Is in one test for sprawling you test the "collapsing distance" to remove some objects being long but so thin that they disappear ?
There was a problem hiding this comment.
would not it be more reasonable to have these tests in gridIndicatorsTest since the methods belong to the gridIndicator class ?
There was a problem hiding this comment.
This class will be removed
| * @param datasource connection to the database | ||
| * @param grid_indicators a grid that contains for each cell the LCZ_PRIMARY | ||
| * @param nb_levels number of aggregate levels. Default is 1 |
| void multiscaleLCZGridTest() { | ||
| //Data for test | ||
| h2GIS.execute(""" | ||
| --Grid values |
| @@ -225,4 +227,128 @@ class SpatialUnitsTests { | |||
| def countRows = postGIS.firstRow "select count(*) as numberOfRows from $outputTable" | |||
| assert 100 == countRows.numberOfRows | |||
| } | |||
|
|
|||
| @Test | |||
| void sprawlAreasTest1() { | |||
This PR proposes new indicators to investigate the characteristics of a territory: spatial distribution of objects, measurement of heterogeneity, neighborhood analyses, distribution of objects according to geographical directions, distances between objects...
Below some outputs on Dijon
Extract sprawl areas
Inside distances to sprawl areas
Outside distances to sprawl areas
Inside cool LCZ distances to sprawl areas
Multiscale grid with adajacent cell values (3*3)
Sprawl surface by compass direction
