[GRASS-user] Not matching categories/line IDs

Johannes_Radinger · September 13, 2018, 8:50am

Hi all,

I am a little bit puzzled about the actual number of lines/categories of a specific vector map (river network). As discussed yesterday, the output of v.category with the report option provides a column ‘count’ which is a feature count. Running v.category option=report on my vector provides me following:

v.category input=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread option=report
Layer/table: 1/Ebro_river_vector_poly_clean_tmp
type count min max
point 0 0 0
line 804 1 1053

…

This would indicate that I have 804 line features. However, using the tool v.info on the same vector I get following different result regarding the number of lines:

v.info map=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread
±---------------------------------------------------------------------------+
| Name: Ebro_river_vector_poly_clean_tmp |
…
| Number of points: 0 Number of centroids: 0 |
| Number of lines: 799 Number of boundaries: 0 |
…

And even more puzzling, I then added a table to this vector map and the output of v.db.addtable is:

v.db.addtable map=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread
Reading features…
Updating database…
803 categories read from vector map (layer 1)
803 categories read from vector map don’t exist in selection from table
803 records updated/inserted (layer 1)

So, is it 804,799 or 803??

The map I am using is a polyline river network (cat=first) that has gone trough the tool v.clean with the options: break,rmdupl,rmline,rmsa,rmdangle. The number of lines and cats before the cleaning tool is 805 (consistent output among all tools).

So I am wondering what happens during cleaning and how to get a clean map with a corresponding table that has entries for each cat, and were each cat represents one line feature (i.e. that the number of cats/line features is consistent over the multiple tools).

Please let me know if I should share the vector map (and how, which format).

Cheers,
Johannes

Markus_Metz · September 13, 2018, 8:26pm

On Thu, Sep 13, 2018 at 10:50 AM Johannes Radinger <johannesradinger@gmail.com> wrote:

Hi all,

I am a little bit puzzled about the actual number of lines/categories of a specific vector map (river network). As discussed yesterday, the output of v.category with the report option provides a column ‘count’ which is a feature count. Running v.category option=report on my vector provides me following:

v.category input=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread option=report
Layer/table: 1/Ebro_river_vector_poly_clean_tmp
type count min max
point 0 0 0
line 804 1 1053
…

This would indicate that I have 804 line features. However, using the tool v.info on the same vector I get following different result regarding the number of lines:
v.info map=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread
±---------------------------------------------------------------------------+
| Name: Ebro_river_vector_poly_clean_tmp |
…
| Number of points: 0 Number of centroids: 0 |
| Number of lines: 799 Number of boundaries: 0 |
…

And even more puzzling, I then added a table to this vector map and the output of v.db.addtable is:
v.db.addtable map=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread
Reading features…
Updating database…
803 categories read from vector map (layer 1)
803 categories read from vector map don’t exist in selection from table

803 records updated/inserted (layer 1)

If there are 804 features, 2 features must have the same category value, resulting in 803 updates to the database

So, is it 804,799 or 803??

The output of v.info and v.category option=report must be identical with regard to the count of features.

The map I am using is a polyline river network (cat=first) that has gone trough the tool v.clean with the options: break,rmdupl,rmline,rmsa,rmdangle. The number of lines and cats before the cleaning tool is 805 (consistent output among all tools).

So I am wondering what happens during cleaning and how to get a clean map with a corresponding table that has entries for each cat, and were each cat represents one line feature (i.e. that the number of cats/line features is consistent over the multiple tools).

Please let me know if I should share the vector map (and how, which format).

Ideal would be a stripped down GRASS mapset with only this vector Ebro_river_vector_poly_clean_tmp. Exporting the vector would change the geometry representation and it would no longer be possible to reproduce the problem.

Markus M

Johannes_Radinger · September 14, 2018, 6:53am

Hi Markus,
Hi all,

I created a zipped version of a test location including one PERMANENT mapset that includes the vector map that causes these inconsistencies (Ebro_river_vector_poly_clean_tmp). The test location can be downloaded from here: https://we.tl/t-eSUZqfP0TQ. I also included the initial vector map before running the v.clean tool (Ebro_river_vector_poly_tmp). More specifically, I am using following code:

grass.run_command(“v.clean”,
overwrite=True,
flags=“c”,
input=“Ebro_river_vector_poly_tmp”,
output=“Ebro_river_vector_poly_clean_tmp”,
tool=“break,rmdupl,rmline,rmsa,rmdangle”,
threshold=“0,0,0,0,0.001”)

Maybe this helps to find out what happens here.

/Johannes

On Thu, Sep 13, 2018 at 10:26 PM Markus Metz <markus.metz.giswork@gmail.com> wrote:

On Thu, Sep 13, 2018 at 10:50 AM Johannes Radinger <johannesradinger@gmail.com> wrote:

Hi all,

I am a little bit puzzled about the actual number of lines/categories of a specific vector map (river network). As discussed yesterday, the output of v.category with the report option provides a column ‘count’ which is a feature count. Running v.category option=report on my vector provides me following:

v.category input=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread option=report
Layer/table: 1/Ebro_river_vector_poly_clean_tmp
type count min max
point 0 0 0
line 804 1 1053
…

This would indicate that I have 804 line features. However, using the tool v.info on the same vector I get following different result regarding the number of lines:
v.info map=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread
±---------------------------------------------------------------------------+
| Name: Ebro_river_vector_poly_clean_tmp |
…
| Number of points: 0 Number of centroids: 0 |
| Number of lines: 799 Number of boundaries: 0 |
…

And even more puzzling, I then added a table to this vector map and the output of v.db.addtable is:
v.db.addtable map=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread
Reading features…
Updating database…
803 categories read from vector map (layer 1)
803 categories read from vector map don’t exist in selection from table

803 records updated/inserted (layer 1)

If there are 804 features, 2 features must have the same category value, resulting in 803 updates to the database

So, is it 804,799 or 803??

The output of v.info and v.category option=report must be identical with regard to the count of features.

The map I am using is a polyline river network (cat=first) that has gone trough the tool v.clean with the options: break,rmdupl,rmline,rmsa,rmdangle. The number of lines and cats before the cleaning tool is 805 (consistent output among all tools).

So I am wondering what happens during cleaning and how to get a clean map with a corresponding table that has entries for each cat, and were each cat represents one line feature (i.e. that the number of cats/line features is consistent over the multiple tools).

Please let me know if I should share the vector map (and how, which format).

Ideal would be a stripped down GRASS mapset with only this vector Ebro_river_vector_poly_clean_tmp. Exporting the vector would change the geometry representation and it would no longer be possible to reproduce the problem.

Markus M

Markus_Metz · September 14, 2018, 8:29am

Apparently the count in v.category option=report is the total number of categories per type and layer. If you have e.g. two lines with categories

line1: 1
line2: 1,2

there are two lines and three categories. This needs to be explained in the manual.

Markus M

On Fri, Sep 14, 2018 at 8:53 AM Johannes Radinger <johannesradinger@gmail.com> wrote:

Hi Markus,
Hi all,

I created a zipped version of a test location including one PERMANENT mapset that includes the vector map that causes these inconsistencies (Ebro_river_vector_poly_clean_tmp). The test location can be downloaded from here: https://we.tl/t-eSUZqfP0TQ. I also included the initial vector map before running the v.clean tool (Ebro_river_vector_poly_tmp). More specifically, I am using following code:

grass.run_command(“v.clean”,
overwrite=True,
flags=“c”,
input=“Ebro_river_vector_poly_tmp”,
output=“Ebro_river_vector_poly_clean_tmp”,
tool=“break,rmdupl,rmline,rmsa,rmdangle”,
threshold=“0,0,0,0,0.001”)

Maybe this helps to find out what happens here.

/Johannes

On Thu, Sep 13, 2018 at 10:26 PM Markus Metz <markus.metz.giswork@gmail.com> wrote:

On Thu, Sep 13, 2018 at 10:50 AM Johannes Radinger <johannesradinger@gmail.com> wrote:

Hi all,

I am a little bit puzzled about the actual number of lines/categories of a specific vector map (river network). As discussed yesterday, the output of v.category with the report option provides a column ‘count’ which is a feature count. Running v.category option=report on my vector provides me following:

v.category input=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread option=report
Layer/table: 1/Ebro_river_vector_poly_clean_tmp
type count min max
point 0 0 0
line 804 1 1053
…

This would indicate that I have 804 line features. However, using the tool v.info on the same vector I get following different result regarding the number of lines:
v.info map=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread
±---------------------------------------------------------------------------+
| Name: Ebro_river_vector_poly_clean_tmp |
…
| Number of points: 0 Number of centroids: 0 |
| Number of lines: 799 Number of boundaries: 0 |
…

And even more puzzling, I then added a table to this vector map and the output of v.db.addtable is:
v.db.addtable map=Ebro_river_vector_poly_clean_tmp@Ebro2_alien_spread
Reading features…
Updating database…
803 categories read from vector map (layer 1)
803 categories read from vector map don’t exist in selection from table

803 records updated/inserted (layer 1)

If there are 804 features, 2 features must have the same category value, resulting in 803 updates to the database

So, is it 804,799 or 803??

The output of v.info and v.category option=report must be identical with regard to the count of features.

The map I am using is a polyline river network (cat=first) that has gone trough the tool v.clean with the options: break,rmdupl,rmline,rmsa,rmdangle. The number of lines and cats before the cleaning tool is 805 (consistent output among all tools).

So I am wondering what happens during cleaning and how to get a clean map with a corresponding table that has entries for each cat, and were each cat represents one line feature (i.e. that the number of cats/line features is consistent over the multiple tools).

Please let me know if I should share the vector map (and how, which format).

Ideal would be a stripped down GRASS mapset with only this vector Ebro_river_vector_poly_clean_tmp. Exporting the vector would change the geometry representation and it would no longer be possible to reproduce the problem.

Markus M