README.md

In MirzaCengic/LUpak:

LUpak package

This is a repo for package that facilitates analysis for the agricultural conversion project.

To install the package run:

devtools::install_github("mirzacengic/LUpak")

Made for personal use only!

Functions currenly added (without particular order):

For modeling:

• variable_importance() - calculate importance of predictor variables.

• get_rasters() - load predictors for specific region.

• load_PA()/load_PA2() - load response variable data (needs to be fixed)

• format_data() - prepare data for modeling (return training and evaluation data). There are some redundant VIF stuff still here.

• vif_stepwise()/vif_select_vars() - calculate VIF by stepwise elimination procedure.

• evaluate_model()/get_evaluations() - Get model assessment results (AUC/TSS).

• fit_model() - Fit linear model to the data.

• categorize_rasters() - Create dummy variable rasters for categorical raster variables. Wraps around raster::layerize(). Needs major overhaul!

For data preparation:

• get_change_raster() - Get differences between two rasters. Optimized version! Might have potential to be generalized for other people to use.

• get_change_raster_GDAL() - alternative with GDAL and parallel implementation.

• rarify_points() - Rarify points according to the raster grid. Optimized version!

• get_absences() - Create absences for agricultural modeling.

MirzaCengic/LUpak documentation built on July 18, 2019, 3:06 a.m.