v0.9.0

PopGen v0.9.0

⚠️⚠️ Breaking Changes

• PopGenCore.jl compat restricted to latest version (very breaking changes)

• kinship methods

  • returns NamedMatrix if no bootstrapping
  • returns DataFrame if bootstrapping

• pairwisefst methods use method = Name instead of method = "Name"

  • fewer keystrokes
  • autocompletion!

✨✨ New Things

• new clustering methods wrapped in generic cluster function:
  • kmedoids
  • hierarchical clustering
  • fuzzy c-means
  • dbscan
• write to Baypass format files
• read/write PLINK (ped)
• t-SNE analysis

⚡⚡ Improvements

• kinship
  • substantially improved speed
  • substantially decreased RAM usage (orders of magnitude)
• pairwisefst
  • better API
  • faster and fewer allocations
• better biallelic conditionals
• a lot of internal functions have been optimized for speed and fewer allocations (PopGenCore.jl)

🐛🐛 Bug fixes

• absolute path eliding in file reading info
• a bunch more I forgot to write down

Diff since v0.8.0

Merged pull requests:

• add more clustering methods (#99) (@pdimens)
• add cluster() wrapper (#100) (@pdimens)