Stream the CANdid dataset into stats¶
Goal¶
Pipe a remote CANdid replay stream directly into the local analysis commands, with no temporary file on disk.
Prerequisites¶
- CANarchy installed and network access to the CANdid dataset.
- The CANdid catalogue is built in; no provider cache refresh is required.
Quick preflight¶
Resolve the replay metadata without opening the stream:
canarchy datasets replay catalog:candid --dry-run --json
List the replayable files in the dataset:
canarchy datasets replay catalog:candid --list-files --json
Bounded replay piped into stats¶
canarchy datasets replay catalog:candid --rate 1000 --max-frames 1000 \
| canarchy stats --file - --json
stats --file - reads candump text from stdin. Combine with
--max-frames to bound the work upfront.
Bounded replay piped into capture-info¶
canarchy datasets replay catalog:candid --rate 1000 --max-frames 1000 \
| canarchy capture-info --file - --json
Select a specific file in the dataset¶
canarchy datasets replay catalog:candid \
--file 2_steering_CAN.log \
--rate 100 \
--max-frames 200 \
--format jsonl