You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Right now the training time is quite long, so we need to focus on a demo that is simple and speedy for a "GEE classification for dummies " type of feeling.
The text was updated successfully, but these errors were encountered:
could this be linked to the PG-viewer SAP that SIN is working on? classification for dummies in GEE works on any pyramid level, so while zoomed out inference occurs at coarsest pyramid level and is super fast...
The bottleneck in our current notebook is not only the size or resolution of the area, but also the fact that we work on a long season.
In the quick version, we can perhaps play around with taking only a few months. Working on a more coarse resolution and benefiting from overviews within the files is also supported, we can do that with a simple 'resample_spatial'.
I believe the random forest inference process is not foreseen in the PG-viewer SAP, so don't think that will be possible.
Limitation of rf_predict() as part of the on-demand viewer (SAP02):
SH does not support rf_predict()
Terrascope with Spark does not read RF models...
Thus on hold for now..
Right now the training time is quite long, so we need to focus on a demo that is simple and speedy for a "GEE classification for dummies " type of feeling.
The text was updated successfully, but these errors were encountered: