search

lutraconsulting / MDAL

Mesh Data Abstraction Library
http://www.mdal.xyz/
MIT License
160 stars 50 forks source link

Bad read of 0 values in SELAFIN file #316

Closed ccoulet closed 3 years ago

ccoulet commented 3 years ago

Hi I noticed sometimes there is hole in the map when I display some Telemac results. After analysis, I discover that the hole are generated where the values on node are 0. When inspect the values, there is no_data indicate.. This could have sense for water depth as sometimes we don't want to map 0 values but this is more problematic for the bottom or the free surface when working on coasts as 0 is a real value... Hope this could be solved in a future version

vcloarec commented 3 years ago

which version?

ccoulet commented 3 years ago

Pi

vcloarec commented 3 years ago

I think this is fixed with last MDAL and QGIS 3.16. Can you try?

ccoulet commented 3 years ago

I should update ;-) I will try and let you know

ccoulet commented 3 years ago

Seams to work with 3.16. But Qgis is slower !!!

vcloarec commented 3 years ago

big mesh? what is slower, the rendering?

ccoulet commented 3 years ago

Not so big mesh. About the slower term, it's just a feeling according to the view of my model just before the update and after. No more investigation done

vcloarec commented 3 years ago

Reading of data has changed with this release, before, all the data was loaded in memory for all time steps. Now, only needed data are loaded for each time step/dataset group, so it is possible, when you change the time step or the dataset group, that rendering cold be slower, but not so much. That will depends on the mesh size and/or the speed of your access to data.

PeterPetrik commented 3 years ago

so if it is fixed in 3.16, we closing this ticket.