Preserving point count when snapping clsutered spatial units?

[adhamenaya]

I have a large dataset with 300,000 observations representing points located at various locations on a network. To improve efficiency, I grouped or clustered these points into about 300 spatial units then I used network.snapobservations to snap these new observations to the network.

Now, I want to analyze the spatial patterns of these new spatial units using the network.Moran function. However, I still want to consider the original count of points (300,000) when performing the analysis.

Is there a way to achieve this, i.e., snap the new spatial units to the network while preserving the information about the original 300,000 points, so I can use this information in the networkMoran function?

[jGaboardi]

Hi, @adhamenaya. Unfortunately, there is no way to perform this. All observations should be taking into consideration at one time.

[adhamenaya]

@jGaboardi thank you for your reply. Any recommendations to optimize the use of the network.snapobservations for such large datasets?

[jGaboardi]

@adhamenaya Sorry for not getting back to you sooner. Within spaghetti, I don't have any recommendations. However, @knaaptime Might have some insight on if this might be possible with a PySAL adjacent project: geosnap?