Datasets:
data exploration and setup
Initial exploration and beetle dataset setup.
Need to address multiple annotations of the same individual (likely that each was looked at two or three times ---something to check on); at least one had length and width transposed (see plot at end of notebooks/EDA-0-1.ipynb): Used PictureID + individual
to get count of individuals (individualID
)
- This gives 11,104; however, we have 39,064 entries and would expect 2 per individual (or 4 if looked at twice, but that's shy of our individual count for everyone and we observed that individual "A00000046078_10" was annotated 3 times).
Some other inconsistencies/questions:
- There 30 different site IDs, but 43 different field site names.
- 576 NEON_sampleIDs, but 577 PictureIDs.
- MOAB_001.S.20180724.jpg is missing scientific name label (this is also the NEON_sampleID_)
We also need to adjust some column names for better consistency (e.g., ID columns).
Would also like to add the reference image (for elytra length/width & centimeter on scalebar) to the README.