…for the modern composable data pipeline
I’m now a low-level nuts-and-bolts developer a few years removed from day-to-day usage. If you have questions, feedback, or expertise to share there are some great venues!
@paleolimbot on Discord/BlueSky/Fosstodon/GitHubI’ll collect comments/questions in the chat here
Often data is distributed in bulk but only a small portion is needed for an analysis. Most readers will let you provide a bounding box or SQL query to limit the data that has to be dealt with in memory.
When your analysis must consider the interaction between every feature in one layer and every feature in another layer, your analysis can probably be phrased as a spatial join. Look for it! Joins are optimized in sf, databases, and elsewhere, and effectively used spatial indexes.
Some join examples: