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
do you mean updating the values of existing points in the dataset?
In that case, running one or two refinement iterations (depending on how much the values change) should already do the trick.
If you want to add more datapoints to the dataset after the construction of the kNN search-graph, there is currently no option for this and one would need to change the code.
One could think of searching for the nearest neighbors of the new points in the existing graph as a starting point and then simply add the new points on the bottom layer and run a few refinement iterations.
If you want to progressively build the search graph for regularly sized blocks, you could instead go for sharding and build independent graphs for each block of data which are then all queried and the results are merged.
But that code path also does not yet support adding shards after the initial construction.
Ggnn does not support dynamic updating of datasets?
The text was updated successfully, but these errors were encountered: