The BAM guide to the analysis of messy bird data Please sign in for contributing with a session in: https://docs.google.com/spreadsheets/d/1kVoG5cml7AFQYc35NNJmbv5Do9h527ZWPtLm6v2IwJQ/edit#gid=234398717
- Data Structure
- description of overall structure (tables)
- autodoc to summarize contents of database (number of locations, surveys, years * what else might we want to know?)
- Data cleaning/processing
- data I/O, odbc, read/write/load etc
- normalizing data types and structure
- correcting problematic values, imputation
- filter, transform, aggregate, join, and other db verbs
- spatial data processing (vector/raster sp/sf objects)
- Data analysis
- removal, distance sampling, offsets, glm
- tree based methods, gbm et al
- bias *variance trade *off
- N *mixture (multiple and single visits)
- ARU stuff
- model evaluation/validation (AUC, bias/variance, etc)
- hierarchical models
- Intepretation, visualization, prediction
- understanding coefficients
- plotting effect sizes and relationships (conditional/marginal effects)
- looking for missing covariates
- prediction and prediction error
- mapping (projections, color scales, thresholds)
- Survey design (~data collection, but informed by all previous chapters)
- optimal design, gap analysis
- protocol (time, radius)
- Reproducibility
- how to improve reproducibility of our studies?
- Useful Tools
- SpaDES
- unmarked
- Geonetwork
- Rmarkdown
- Github
- GitKraken
- BAM operations and procedures guide
- orientation/welcome (BAM context, seminal BAM papers, ...)
- coauthorship/collaboration on BAM manuscripts, BAM paper proces and timeline
- BAM sharepoint, shared drives, etc.
- Theme Teams / hierarchical working groups (+ dealing with overlap among projects within and among themes)
- Cheat sheet on where to find various types of information
- Outreach and product distribution