Dear flexsdm user, if you are interested in receiving email notifications about modifications made to the package (e.g., new functions, arguments, or vignettes), please fill out this form so we can keep you updated on what is new in flexsdm.
Species distribution modeling has become a standard tool in several research areas such as ecology, conservation biology, biogeography, paleobiogeography, and epidemiology. Species distribution modeling is an area of active research in both theoretical and methodological aspects. One of the most exciting features of flexsdm is its high manipulation and parametrization capacity based on different functions and arguments. These attributes enable users to define a complete or partial modeling workflow specific for a modeling situation (e.g., number of variables, number of records, different algorithms, algorithms tuning, ensemble methods).
The functions of flexsdm package are organized into three major modeling steps
Set tools that prepare modeling input data (e.g., species occurrences thinning, sample pseudo-absences or background points, delimitation of calibration area).
calib_area()Delimit calibration area for constructing species distribution modelscorrect_colinvar()Collinearity reduction on predictorsenv_outliers()Integration of outliers detection methods in the environmental spacepart_random()Data partitioning for training and testing modelspart_sblock()Spatial block cross validationpart_sband()Spatial band cross validationpart_senv()Environmental cross-validationplot_res()Plot different resolutions to be used in part_sblockget_block()Transform a spatial partition layer to the same spatial properties of environmental variablessample_background()Sample background pointssample_pseudoabs()Sampel pseudo-absencesdm_directory()Create directories for saving the outputs of the flexsdmsdm_extract()Extract environmental data based on x and y coordinatesoccfilt_env()Perform environmental filtering on species occurrencesoccfilt_geo()Perform geographical filtering on species occurrencesoccfilt_select()Select filtered occurrences when it was tested with different filtering values
It includes functions related to modeling construction and validation. Several of them can be grouped into fit_*, tune_*, and esm_* family functions. fit_* construct and validate models with default hyper-parameter values. tune_* construct and validate models searching for the best hyper-parameter values combination. esm_ construct and validate Ensemble of Small Models.
sdm_eval()Calculate different model performance metrics
fit_gam()Fit and validate Generalized Additive Modelsfit_gau()Fit and validate Gaussian Process modelsfit_gbm()Fit and validate Generalized Boosted Regression modelsfit_glm()Fit and validate Generalized Linear Modelsfit_max()Fit and validate Maximum Entropy modelsfit_net()Fit and validate Neural Networks modelsfit_raf()Fit and validate Random Forest modelsfit_svm()Fit and validate Support Vector Machine models
tune_gbm()Fit and validate Generalized Boosted Regression models with exploration of hyper-parameterstune_max()Fit and validate Maximum Entropy models with exploration of hyper-parameterstune_net()Fit and validate Neural Networks models with exploration of hyper-parameterstune_raf()Fit and validate Random Forest models with exploration of hyper-parameterstune_svm()Fit and validate Support Vector Machine models with exploration of hyper-parameters
fit_ensemble()Fit and validate ensemble models with different ensemble methods
esm_gam()Fit and validate Generalized Additive Models with Ensemble of Small Model approachesm_gau()Fit and validate Gaussian Process models Models with Ensemble of Small Model approachesm_gbm()Fit and validate Generalized Boosted Regression models with Ensemble of Small Model approachesm_glm()Fit and validate Generalized Linear Models with Ensemble of Small Model approachesm_max()Fit and validate Maximum Entropy models with Ensemble of Small Model approachesm_net()Fit and validate Neural Networks models with Ensemble of Small Model approachesm_svm()Fit and validate Support Vector Machine models with Ensemble of Small Model approach
Tools related to models’ geographical predictions, evaluation, and correction.
sdm_predict()Spatial predictions of individual and ensemble modelsdm_summarize()Merge model performance tablesinterp()Raster interpolation between two time periodsextra_eval()Measure model extrapolationextra_truncate()Constraint suitability values under a given extrapolation valuemsdm_priori()Create spatial predictor variables to reduce overprediction of species distribution modelsmsdm_posteriori()Methods to correct overprediction of species distribution models based on occurrences and suitability patterns.
Useful tools to visually explore models’ geographical and environemtal predictions, model extrapolation, and partial depnendece plot.
p_pdp()Create partial dependence plot(s) to explore the marginal effect of predictors on suitabilityp_bpdp()Create partial dependence surface plot(s) to explore the bivariate marginal effect of predictors on suitabilityp_extra()Graphical exploration of extrapolation or suitability pattern in the environmental and geographical spacedata_pdp()Calculate data to construct partial dependence plotsdata_bpdp()Calculate data to construct partial dependence surface plots
You can install the development version of flexsdm from github
# install.packages("remotes")
# For Windows and Mac OS operating systems
remotes::install_github("sjevelazco/flexsdm")
# For Linux operating system
remotes::install_github("sjevelazco/flexsdm@HEAD")See the package website (https://sjevelazco.github.io/flexsdm/) for functions explanation and vignettes.
Velazco, S.J.E., Rose, M.B., Andrade, A.F.A., Minoli, I., Franklin, J. (2022). flexsdm: An R package for supporting a comprehensive and flexible species distribution modelling workflow. Methods in Ecology and Evolution, 13(8) 1661–1669. https://doi.org/10.1111/2041-210X.13874
Test the package and give us your feedback here or send an e-mail to [email protected]{.email}.