calibrate_sicktree_model_multi_tile: Calibrate vegetation distribution models

In pieterbeck/CanHeMonR: Tools for canopy health monitoring through image analysis

Description

For each class in .shp polygon file, calibrate a distribution model using a raster brick as predictors

Usage

calibrate_sicktree_model_multi_tile(vuln_classes = "ALL", training_df,
  model_outp_dir)

Arguments

vuln_classes	A character vector of the classes you want to model. The should be presented in the column 'class' of training_df.
training_df	data.frame, with in the column 'pres' 1/0 to indicate presence absence, then covariate columns, and a colum 'class' groupin grows by the land-cover class the data was sampled for. This df is typically generated by sample_for_sicktree_model_multi_tile
model_outp_dir	The folder and filename prefix to save the model objects to

Value

Saves class-specific distribution models, using a data frame created from training points and covariate images

Note

Run in 32-bit R installation. Do you need a 'require(rJava)?'. Implement optional parallel

Examples

## Not run: 
#calibrate the model with 10 absences per tile
model_dir <- '//ies.jrc.it/h03/CANHEMON/H03_CANHEMON/Imagery/Portugal/DMC/ortophotos_22122016/classification_temp/'
testmod100 <- calibrate_sicktree_model_multi_tile(vuln_classes='ALL',
training_df = readRDS(file.path(model_dir, 'maxent_samplemaxent_training_dfs_100_samples.rdsdata')),
                                                  model_outp_dir = paste0(model_dir,'samp100_'))

#calibrate the model with 10 absences per tile
model_dir <- '//ies.jrc.it/h03/CANHEMON/H03_CANHEMON/Imagery/Portugal/DMC/ortophotos_22122016/classification_temp/'
testmod10 <- calibrate_sicktree_model_multi_tile(vuln_classes='ALL',
training_df = readRDS(file.path(model_dir, 'maxent_samplemaxent_training_dfs_10_samples.rdsdata')),
model_outp_dir = paste0(model_dir,'samp10_')_

## End(Not run)

pieterbeck/CanHeMonR documentation built on May 25, 2019, 7:11 a.m.