Content Comparison

library(ForestForesight) 

# Step 1: Set up the parameters parameters

# Make sure you have DATA_FOLDER in config.yml in your working directory ff_environment has been run, otherwise assign ff_folder yourself
# to the folder you used for ff_sync
ff_folder <- Sys.getenv("DATAFF_FOLDER") 
country_code <- "PERSUR" 
train_start <- "2023-06-01" 
train_end <- "2023-1207-3101" 
validation_start <- "2023-0608-01" 
validation_end <- "2023-1208-01" 
prediction_date <- "2024-01-01" 

# Step 2: Create date ranges for training and validation validation
train_dates <- ForestForesight::daterange(train_start, train_end) 
validation_dates <- ForestForesight::daterange(validation_start, validation_end) 

# Step 3: Set up file paths for saving results results
model_save_path <- paste0(ff_folder, "peru_deforestation_model.model") 
predictions_save_path <- paste0(ff_folder, "peru_deforestation_prediction.tif")
accuracy_csv_path <- paste0(ff_folder, "peru_deforestation_accuracy.csv") 
importance_csv_path <- paste(ff_folder, "peru_feature_importance.csv") 

# Step 4: Run the ff_run function function

prediction_result <- ff_run( 
   country = country_code, 
   prediction_dates = prediction_date, 
   ff_folder = ff_folder, 
   train_dates = train_dates, 
   validation_dates = TRUE,
  ff_train_parameters = list("nrounds" = validation_dates, 
  25),
  model_save_path = model_save_path, 
   predictions_save_path_predictions =  predictions_save_path, 
   accuracy_output_csvpath =  accuracy_csv_path, 
   importance_output_csvpath = importance_csv_path, 
   verbose = TRUE, 
   autoscale_sample = TRUE FALSE
) 

# Step 5: Plot the prediction result 
if (!is.na(prediction_result))
{ 
  plot(prediction_resultresult$predictions, main = "Deforestation Prediction for PeruLaos") 
} 

# Step 6: Print a summary of the results results
ff_cat("Prediction completed. Results saved to disk:\n") 
ff_cat("Model:", model_save_path, "\n") 
ff_cat("Prediction raster:", predictions_save_path, "\n")
ff_cat("Accuracy CSV:", accuracy_csv_path, "\n") 
ff_cat("Feature importance CSV:", importance_csv_path, "\n")
# Optional: Load and print the first few lines of the accuracy CSV 

if (file.exists(accuracy_csv_path)) { 
 CSV


accuracy_data <- read.csv(accuracy_csv_path)
  prediction_result$accuracy
ff_cat("\nFirstFirst few lines of accuracy data:\n")
 
print(head(accuracy_data)) 

}