R/rbms_index_modelling.r
fit_gam.Rdfit_gam fit a Generalized Additive Model to butterfly count data along a temporal variable and accounting for site effect when multiple are available.
fit_gam(
dataset_y,
NbrSample = NULL,
GamFamily = "poisson",
MaxTrial = 4,
SpeedGam = TRUE,
OptiGam = TRUE,
TimeUnit = "d",
MultiVisit = "mean",
weekday = 3,
mod_form = NULL,
tp_col = NULL,
verbose = TRUE,
...
)data.table Filtered butterfly counts for species x over year y over all sites.
integer Inherited from flight_curve, default=100.
string Inherited from flight_curve, default='poisson', but can be 'nb' or 'quasipoisson'.
integer Inherited from flight_curve, default=3.
logical Set if the bam method should be used, default instead of the default gam method.
character The time-step for which the spline should be computed, 'd' day or 'w' week.
string Function to apply for summarising multiple counts within a time unit, 'max' or 'mean' (default).
Integer for selected day of the week for weekly summary, default is 3 (Wednesday). [1-7] where 1 = Monday.
string with formula to be passed to the gam model, default null.
string or vector of string with additional variable used in the gam model, default null.
a logical indicating if some “progress report” should be given.
Additional parameters passed to gam or bam function from the gam package.
A list with three objects, i) **f_curve**: a data.table with the flight curve f_curve with expected relative abundance, normalize to sum to one over a full season,
ii) **f_model**: the resulting gam model f_model fitted on the count data and iii) **f_data**: a data.table with the data used to fit the GAM model. This is provide for one year 'y'.