Datasets:
stim_order.parquet
Thanks for publishing your dataset! I’m trying to preprocess the raw data using the code provided in your GitHub repository. According to the preprocessing.py script, there should be a stim_order.parquet file, but I wasn’t able to locate it. Am I missing something?
Same question
I looked through the revision history, and it seems like they removed the stim_order when they cleaned up the metadata on commit 65613cf0c298910c3b9ffa20096a50cf0148a6d3: "Remove autogenerated metadata (.parquet, .pkl)".
If you're using the HuggingFace hub package, you can download them for a specific sub with the command:
huggingface_hub.snapshot_download("Alljoined/Alljoined-1.6M", allow_patterns=f"*{SUB:02}*/stim_order.parquet", repo_type="dataset", local_dir=OUTPUT_DIR, revision="bd55487200c4c712dac9be2ee6fc8f593095b454")
I'm not sure if this is the cleanest way to go about it, or if there is an official method supported.
Thanks for pointing this out! We have added back the stim_order.parquet files in PR #6, which just merged, so they are now included again in the latest version of the dataset.