qgis:dissolve - "dissolve all" does not dissolve attributes - Githubissues

qgib commented 11 years ago

Author Name: Pedro Venâncio (Pedro Venâncio) Original Redmine Issue: 6918 Affected QGIS version: 2.4.0 Redmine category:processing/qgis Assignee: Victor Olaya

New description:

the qgis:dissolve tool ("dissolve all" option) does not dissolve the attributes, just the geometries. The original tool in the vector menu is ok

Old description:

Hi,

'Dissolve all' option is not taking effect. The result of the following commands is exactly the same.

sextante.runalg("qgis:dissolve","/home/pedro/Comunicacoes/Encontro_SMPC/Dados_Geograficos/Cont_AAd_CAOP20121/Cont_AAd_CAOP20121_dlxhgmil_3_Pinhel.shp",False,True,"Freguesia",None)

sextante.runalg("qgis:dissolve","/home/pedro/Comunicacoes/Encontro_SMPC/Dados_Geograficos/Cont_AAd_CAOP20121/Cont_AAd_CAOP20121_dlxhgmil_3_Pinhel.shp",False,False,"Freguesia",None)

I'm using QGIS master (today's Github) on Linux.

dissolve_all.jpg (Pedro Venâncio)
dissolve_all_2.png (Pedro Venâncio)

qgib commented 11 years ago

Author Name: Giovanni Manghi (@gioman)

this used to work, isn't it?

qgib commented 11 years ago

Author Name: Pedro Venâncio (Pedro Venâncio)

Hi Giovanni,

Yes, at this time I can not tell what the last 'version' that worked, but definitely it used to work.

qgib commented 11 years ago

Author Name: Giovanni Manghi (@gioman)

priority_id was changed from Normal to Severe/Regression

qgib commented 11 years ago

Author Name: Victor Olaya (@volaya)

Fixed. The problem was that it was using the value of the other true/false field for the "use all" field. And, checking in github, the history of that file in SEXTANTE says it has been like that since the very first day...so I guess it never worked right :-)

You will see that now there is no "use selected" option. I am removing them and adding them as a global option, so algorithms are more compact and can be written with less code. This algorithm is adapted already, but I still have to work on the other fTools ones.

status_id was changed from Open to Resolved

qgib commented 11 years ago

Author Name: Pedro Venâncio (Pedro Venâncio)

Hi Victor,

Funny, I almost swore that had seen him working properly! :)

Thank you very much!

qgib commented 11 years ago

Author Name: Pedro Venâncio (Pedro Venâncio)

status_id was changed from Resolved to Closed

qgib commented 11 years ago

Author Name: Anita Graser (@anitagraser)

I'm testing with current nighly and "dissolve all" yes does not dissolve all features.

Also, it makes no sense that I have to specify a "Unique ID field" if I want to dissolve all. (In model builder GUI)

status_id was changed from Closed to Reopened

qgib commented 11 years ago

Author Name: Anita Graser (@anitagraser)

It's really odd, because if I run Dissolve directly in the toolbox and select "dissolve all" NO, then it does dissolve all ...

qgib commented 11 years ago

Author Name: Pedro Venâncio (Pedro Venâncio)

Hi,

Anita Graser wrote:

It's really odd, because if I run Dissolve directly in the toolbox and select "dissolve all" NO, then it does dissolve all ...

I can not confirm this. With dissolve all set to NO, it works ok. But with dissolve all set to YES, the geometries appears dissolved, but the attribute table has the same number of registers than the original layer minus 1, and the attributes of the first register appears repeated n-1 times. See the image attached.

Another issue is to dissolve big datasets. In case, I tried with a shapefile with 8612 records, and after 2 hours, there is no result.

5627 was configured as dissolve_all.jpg

dissolve_all.jpg (Pedro Venâncio)

qgib commented 11 years ago

Author Name: Giovanni Manghi (@gioman)

Another issue is to dissolve big datasets. In case, I tried with a shapefile with 8612 records, and after 2 hours, there is no result.

this is known issue, the qgis dissolve is very bad and until its code will be revised it basically necessary to use GRASS or SAGA... and with sextante is easy :)

qgib commented 11 years ago

Author Name: Giovanni Manghi (@gioman)

Anita Graser wrote:

It's really odd, because if I run Dissolve directly in the toolbox and select "dissolve all" NO, then it does dissolve all ...

I also cannot confirm this, tested with different datasets on both Linux and Windows. Please reopen if necessary.

status_id was changed from Reopened to Closed

qgib commented 11 years ago

Author Name: Salvatore Larosa (@slarosa)

Giovanni Manghi wrote:

I also cannot confirm this, tested with different datasets on both Linux and Windows. Please reopen if necessary.

do rows in attributes table are dissolved too ?

qgib commented 11 years ago

Author Name: Giovanni Manghi (@gioman)

do rows in attributes table are dissolved too ?

no :(

subject was changed from qgis:dissolve - Dissolve all not working to qgis:dissolve - "dissolve all" does not dissolve attributes
status_id was changed from Closed to Reopened
priority_id was changed from Severe/Regression to High

qgib commented 11 years ago

Author Name: Filipe Dias (@fsdias)

Still true?

qgib commented 11 years ago

Author Name: Giovanni Manghi (@gioman)

Filipe Dias wrote:

Still true?

yes

qgib commented 11 years ago

Author Name: Pedro Venâncio (Pedro Venâncio)

Yes Filipe.

6361 was configured as dissolve_all_2.png

dissolve_all_2.png (Pedro Venâncio)

qgib commented 10 years ago

Author Name: Giovanni Manghi (@gioman)

version was configured as 2.4.0
project_id was changed from 78 to 17
category_id removed 61
crashes_corrupts_data was configured as 0

qgib commented 10 years ago

Author Name: Giovanni Manghi (@gioman)

category_id was configured as Processing/QGIS

qgib commented 9 years ago

Author Name: Pedro Venâncio (Pedro Venâncio)

I think this was fixed in this https://github.com/qgis/QGIS/blob/116089e6ec0bd9bfd15395206bd6e880561da859/python/plugins/processing/algs/qgis/Dissolve.py

Reopen if necessary.

status_id was changed from Reopened to Closed

qgis / QGIS

qgis:dissolve - "dissolve all" does not dissolve attributes #16037