Makes file with media manifest content IDs not in predicted catalog for MD5 checks.

Notebooks to explore mismatch between catalog files and media manifest they should represent.

We may have a content/page ID mismatch.

Update file paths to match data move and read in EOL content/page ID cols as ints.

Make manifests of missing pages

Explore content of newer media manifest (Dec. 6 download in EOL repo) and compare earlier media manifest.

Merge and organize git lfs files.

v3.3 update visualizations (#4)

Compare content IDs from Media Cargo (all downloaded images).

Merge branch 'main' of https://huggingface.co/datasets/imageomics/ToL-EDA into pr/5

v3.3 Dataset Diversity Overlap Analysis (#6)

Add data files with missing content IDs, one with taxa info too.

Look at some of the taxa distribution of images missing from the media manifest.

Confirmed that content IDs that don't match to media manifest are in both TreeOfLife-10M and Rare Species datasets.

Check media manifests from July 6 and 26 to confirm coverage of all listed images; 113K missing link to media manifest.

Remove train_small duplication and add file indicating images missing owner of licenses.

Analysis of EOL license file, missing content and some missing licenses.

Generate csv to store license files for all predicted images in dataset from EOL. This will include Rare Species as well.

v3.3 EDA update (#3)

v3.2 EDA Update (#2)

Updated EDA (#1)

Updated count information to include all values that should show up as warnings when running check_taxa.

Generated CSV for testing check_taxa script (auto-check for filled-in taxonomic hierarchy).

Add notebooks analyzing the dataset contents, taxa distributions, etc. and generating visualizations.