-
Delete metrics.parquet
-
Simplify data into three source tables and add ZIP conversion script
-
Simplify README
-
Import all records from ten source ZIP archives without filtering
-
Keep the dataset description independent of its size
-
Delete cycling_sample.parquet
-
Use stable table names and general data checks
-
Delete source_samples.csv
-
Delete source_metadata.json
-
Delete source_context.json
-
Delete build-requirements.txt
-
Delete build_metadata.py
-
Remove the old measurement-only query script
-
Add complete activity metadata and linked SQL tables to the example
-
Write dataset instructions in plain English
-
Use viewer-compatible numeric types and add sample validation
-
Add small cycling Parquet sample and DuckDB SQL example
-
initial commit
