Skip to contents

flight_curve_multi Compute flight curves across multiple years simultaneously using a single BAM model with factor-smooth interactions. This is more efficient than fitting separate models per year and can provide better estimates for years with sparse data.

Source: R/fit_bam_multi.r
flight_curve_multi.Rd

flight_curve_multi Compute flight curves across multiple years simultaneously using a single BAM model with factor-smooth interactions. This is more efficient than fitting separate models per year and can provide better estimates for years with sparse data.

Usage

flight_curve_multi(
  ts_season_count,
  NbrSample = 100,
  MinVisit = 3,
  MinOccur = 2,
  MinNbrSite = 1,
  MaxTrial = 3,
  GamFamily = "poisson",
  CompltSeason = TRUE,
  SelectYear = NULL,
  TimeUnit = "d",
  MultiVisit = "mean",
  weekday = 3,
  tp_col = NULL,
  smooth_basis = "cr",
  k_value = -1,
  KeepModel = TRUE,
  KeepModelData = TRUE,
  verbose = TRUE,
  ...
)

Arguments

ts_season_count

data.table Time-series of counts and season information returned by ts_monit_count_site

NbrSample

integer Maximum number of sites to use per year, default=100.

MinVisit

integer Minimum number of visits required for a site to be included, default=3.

MinOccur

integer Minimum number of positive records required per year in a site, default=2.

MinNbrSite

integer Minimum number of sites required per year, default=1.

MaxTrial

integer Maximum number of trials for model convergence, default=3.

GamFamily

string Distribution for GAM, default='poisson', but can be 'nb' or 'quasipoisson'.

CompltSeason

logical Restrict to years where complete season was sampled, default=TRUE.

SelectYear

integer Vector of specific years to include, default=NULL (use all).

TimeUnit

character Time-step for the spline, 'd' for day or 'w' for week.

MultiVisit

string Function for summarising multiple counts within a time unit, 'max' or 'mean' (default).

weekday

Integer for selected day of the week for weekly summary, default is 3 (Wednesday). [1-7] where 1 = Monday.

tp_col

string Name of temporal variable, default NULL (auto-determined from TimeUnit).

smooth_basis

string Basis for smooth term, default "cr". Options: "tp", "ps", "cr".

k_value

integer Basis dimension, default=-1 (automatic).

KeepModel

logical Keep model output, default=TRUE.

KeepModelData

logical Keep data used for fitting, default=TRUE.

verbose

logical Show progress reports, default=TRUE.

...

Additional parameters passed to bam function.

Value

A list with three objects: i) **pheno**: flight curves with expected relative abundance (NM) across all years, ii) **model**: the bam model object fitted across all years, iii) **data**: the data used to fit the model (if KeepModelData=TRUE).

Details

This function uses mgcv::bam with factor-smooth interactions which is particularly useful for: - Large datasets with many years - Years with varying data quality/quantity - Sharing information across years while allowing year-specific patterns

See also

fit_bam_multi, flight_curve, bam

Author

Reto Schmucki - retoshm@ceh.ac.uk