Colin Ho
Colin Ho
# Daft Roadmaps Our roadmaps are on our discussions page. You can find a complete list of all Daft roadmaps here https://github.com/Eventual-Inc/Daft/discussions/categories/roadmaps ### Ongoing Projects * **https://github.com/Eventual-Inc/Daft/discussions/5155**: Revamp our observability...
## Changes Made Adds logic to elide shuffles for either side of a hash join when input num partitions != num partitions for join or input is not hash partitioned...
## Changes Made Raises error in situations where actors for actor UDFs (i.e. concurrency udfs) cannot be scheduled due to resource constraints. This is implemented via timeout (default 60s) on...
## Changes Made ## Related Issues ## Checklist - [ ] Documented in API Docs (if applicable) - [ ] Documented in User Guide (if applicable) - [ ] If...
### Is your feature request related to a problem? There aren't enough tutorials / examples on using Daft for AI / ML workloads. ### Describe the solution you'd like Some...
### Is your feature request related to a problem? `count(*)` or `df.count()` is not as performant as can be. ### Describe the solution you'd like We should be able to...
## Changes Made ## Related Issues ## Checklist - [ ] Documented in API Docs (if applicable) - [ ] Documented in User Guide (if applicable) - [ ] If...
## Changes Made ## Related Issues ## Checklist - [ ] Documented in API Docs (if applicable) - [ ] Documented in User Guide (if applicable) - [ ] If...
## Changes Made Initialize the subprocess for udfworker lazily upon first batch. This is so that we don't unnecessarily spawn a subprocess if it's not going to execute on any...
### Is your feature request related to a problem? Support python 3.14 (and 3.14t). Some known issues are related to type checking https://github.com/Eventual-Inc/Daft/issues/5536 and https://github.com/Eventual-Inc/Daft/issues/5538 ### Describe the solution you'd...