train_dl: Train deep learning SDM
In boyanangelov/sdmbench: Benchmark Species Distribution Models

Description Usage Arguments Value Examples

Train deep learning SDM

1	train_dl(input_data)

input_data

A dataframe containing parsed and processed deep learning data.

A list containing the trained deep learning model object and a history object for performance evaluation.

## Not run: 
# download benchmarking data
benchmarking_data <- get_benchmarking_data("Lynx lynx",
                                           limit = 1500)

# transform benchmarking data into a format suitable for deep learning
# if you have previously used a partitioning method you should specify it here
benchmarking_data_dl <- prepare_dl_data(input_data = benchmarking_data$df_data,
                                       partitioning_type = "default")

# perform sanity check on the transformed dataset
# for the training set
head(benchmarking_data_dl$train_tbl)
table(benchmarking_data_dl$y_train_vec)

# for the test set
head(benchmarking_data_dl$test_tbl)
table(benchmarking_data_dl$y_test_vec)

# train neural network
keras_results <- train_dl(benchmarking_data_dl)

# inspect results
keras_results$history

# you can also plot them
plot(keras_results$history)

## End(Not run)