AlexTasks7 19 2018

1. Transform preprocessing notebook (notebooks/brain/preprocess/preprocess_cshl_data_v2_neurotrace.ipynb) into a scripts with argparse parameters. Write test scripts for these scripts.
2. Figure out the format of the files coming from UCSD and how they are translated to jp2000 (uncompressed)
3. Find a simple alternative to Yuncong's file-path-manager. (One idea: a script that asks you to enter directory paths and stores them in a configuration file and sets them into environmental variables). If file exists, just use it unless over-ride.
4. clean up old documentation. put historical links in a separate file.
5. Move out of old test files that are relevant to the files under src.

---

Added by Yoav: Clean up the bucket: s3://MouseBrain-data/.

Separate into:

• unreproducible Data that cannot be reproduced, and is necessary for the functioning of the atlas.
• reproducible Data that can be reproduced: keep only latest version, document how the data can be reproduced.
• Old files that might be needed in the future.

Steps:

1. Compile a list of all directories
2. Mark directories as unreproducible, reproducible, Old
3. Write a (python) script to make the required changes in the buckets (use AWS CLI)