Package index • BirdFlowPipeline

All functions

BatchBirdFlowEvaluator(): Construct a batch evaluator
BatchBirdFlowTrainer(): Construct a batch BirdFlow trainer
TransitionsLoader(): Create a transitions loader
americas_clip: Clipping polygon for the Americas
americas_crs: Custom coordinate reference system for American BirdFlow Models
banding_data_to_linestring(): Function to convert banding data to linestring of origin-destination pairs
banding_raw_file_urls: URLs for downloading raw USGS banding datafiles version 2022-07-14
batch_flux(): Calculate flux for multiple species
batch_modelfit_wrapper(): Function to batch fit models on the cluster, including resubmitting any failed jobs
batch_species(): Fit many species with identical parameters
birdflow_modelfit(): Function to fit BirdFlow model by executing update_hdf.py in Python. Arguments typically passed from batch_modelfit_wrapper() via birdflow_modelfit_args_df()
birdflow_modelfit_args_df(): Create a grid-expanded data.frame of model fit arguments. Output designed to be passed to batch_modelfit_wrapper()
combine_and_save_ground_truth_data(): Combining the multsource data for one species
combine_data_for_all_sp(): Combining the multsource data for all species
combine_together_list(): Make the list of file sets required by process_file_set()
download_banding_files(): Download raw USGS banding files
evaluate(<BatchBirdFlowEvaluator>): Evaluate a set of BirdFlow HDF5 models (S3 method)
evaluate(): Generic evaluation S3
evaluate_model(): Evaluate a single BirdFlow model (default evaluation function)
fit(<BatchBirdFlowTrainer>): Fit a BatchBirdFlowTrainer
fit(): Generic fit S3
format_metrics(): Format metrics for plotting
get_job_efficiency(): get the job efficiency of a completed job from slurm
get_real_track(): Get the BirdFlowRoutes object of the real track
get_transitions(): Get ground-truth transitions for tuning and validation
gpu_ram(): Calculate the GPU RAM needed for a model
identify_best_models(): Identify the best models in grid search
identify_hdf5_model(): Read key parameters from hdf5 modelfit
load_banding_df(): Load banding data
load_data(): Generic loader S3
load_data(<TransitionsLoader>): Load transitions into a TransitionsLoader (S3 method)
load_motus_df(): Load Motus data
make_3d_plot(): 3d plot function
make_flux_index(): Make html index file for flux reports
make_flux_report(): Make flux report
make_new_report(): Make species report
make_species_index(): Function to make an index file for output of batch_species()
model_evaluation_biplot(): PCA biplot hyperparameters evaluation
new_BatchBirdFlowEvaluator(): Initialize a batch evaluator (low-level constructor)
new_BatchBirdFlowTrainer(): Initialize a batch trainer (low-level constructor)
new_transitions_loader(): Initialize a transitions loader (low-level constructor)
plot_grid_search_metrics(): Plot grid search metrics
plot_tuning(): Plot gridsearch performance metrics
preprocess_and_fit(): Process and fit a single model with fixed parameters
preprocess_banding_data_to_rds(): Batch preprocess banding data on the cluster
preprocess_species_wrapper(): Preprocess species and enrich batch parameters
process_file_set(): Function to take a set of raw files and produce per-species RDS
process_flux(): Import a BirdFlow model, calculate flux, and save to an RDS file
quick_visualize_routes(): Quick visualize a spring migration routes simulation from eval_metrics by row number
rank_models(): Rank the models in an eval_metrics data frame
rebuild_model_reports(): Rebuild model report HTML files from existing output directory
refactor_hyperparams(): Find a dist_weight and ent_weight consistent with a given obs_prop and de_ratio
rts_stats(): Calculate route summary statistics
safe_numeric(): Safely return a real value
set_pipeline_params(): Function to set BirdFlowPipeline parameters
split_data(): Generic split S3
split_data(<TransitionsLoader>): Split transitions into training and test sets (S3 method)
stage_model_release(): Copy model file and html model report from pipeline output to release staging folder
taxonomy_crosswalk: data.frame for matching USGS banding taxa to ebirdst taxa
train_test_split(): Train/test split helper for transitions
validate_BatchBirdFlowEvaluator(): Validate a BatchBirdFlowEvaluator object
validate_BatchBirdFlowTrainer(): Validate a BatchBirdFlowTrainer object
validate_TransitionsLoader(): Validate a TransitionsLoader object
validate_split_data(): Validate split data structure