psolymos/bamanalytics

Boreal Avian Modeling Project

Hi there 👋

My name is Péter Sólymos. I am bridging the gap between data and decision making using data analytics, visualization, and infrastructure automation.

• R packages by me: https://psolymos.r-universe.dev/
• Writing about Shiny hosting: https://hosting.analythium.io/
• Edmonton R User Group organizer: https://yegrug.github.io/
• Personal website with publications: https://peter.solymos.org/

bamanalytics's Issues

Prediction

• make it run on WG
• predict current (a base year TBD)
• predict reference: all athropogenic disturbances are 0-ed out (not fire)
• land cover classes agr and disturbed: random non-disturbed in bootstrap (separate for each BCR/prov region unit)

Fix before upscaling

Issues to fix:

• check max of Y (e.g. 67 for CAWA meant 6-7 and should be kept as 6)
• swap climate and disturbance in mods stages
• reduce habitat levels for ROAD:HAB interaction (decid, conif, shrub, wet open, open incl. development), also don't use ROAD:HAB:TREE3 interaction.
• include sqrt(HGT) in models
• se YSD and HGT in same stage and let them compete
• HGT=0 for non-forest and non-wet NALC classes
• EXCLUDE 4x4 (or use landscape level stuff after climate (also based on opticut))
• use new version of offsets