I have a problem with SMART 7.4, at the CCAA level.
Indeed, when I mount and execute a summary query on SMART having in options the selected conservation areas (in row headers) and the type of patrol team or the type of transport (in column headers), and in value the number of patrols or the distance traveled, an error message appears, telling me impossible the query (see image below).
@drew.cronin see screenshots (query and Error logs). the distance query cannot execute for more than 30 days if the group by option selected is one of the area group bys eg “Conservation area boundary” but executes perfectly fine if team or a conservation area is selected.
In my conservation area, for the year 2022, when I run a summary query that takes in the groups by options - the defined administrative area boundaries - and in the values the distance (km) for a period from January 01 to April 30, 2022 (about 100 days), the query does not run and an error message appears.
However, when I run the same query for the period from June to December 2022, the query executes very well.
Apologies for not posting the solution previously (i.e. upgrading).
This is a strange error. Usually when there are issues related to the area filters the shapefile is often to blame but it appears the duration is somehow contributing to the resulting error.
Could I ask that the error log is exported and shared as .txt file, please? We will get someone to look into this. (I’m also not able to load your screenshots Vigny, so if you could re-share, that would be most helpful.
Our developers have been looking into this. Thanks again for bringing this to our attention. It appears we need to make some changes in SMART to avoid these issues but I wanted to provide the explanation offered by our developers:
"both of these errors occurs for some time frames and not others is because the data (geometries) returned in the time is different and it’s the processing of these geometries that is failing.
These sorts of topology errors can occur when coordinates on lines are almost exact but not quite exact. In the first example the startpoint of the first line is almost but not quite the same as the endpoint of the second line:
The only current workaround here is changing the track points (in your patrol data) that are used to generate the query data. But we will be looking into how to avoid this issue in future by updating SMART to ignore more than 10 or 11 digits to avoid this issue entirely. Apologies for the inconvenience caused in the meantime.