API logging library and load H2O.ai model

api-logging-model-load.r

library('logger')
library(h2o)
library('data.table')

h2o.init()

ROOT_DIR <- getwd()

PROJECT_DIR <- 
  'r-api-data-hackers'


MODELS_DIR <- 'models'
API_DIR <- 'api'
LOGS_DIR <- 'logs'

model_path_object <- 
  file.path(ROOT_DIR,
            PROJECT_DIR,
            MODELS_DIR,
            "StackedEnsemble_AllModels_AutoML_20200428_181354")

logging_file_path <- 
  file.path(ROOT_DIR,
            PROJECT_DIR,
            LOGS_DIR,
            "api_predictions.log")

log_appender(appender_file(logging_file_path))
log_layout(layout_glue_colors)
log_threshold(DEBUG)

log_info('Load saved model')
saved_model <- 
  h2o.loadModel(model_path_object)
log_info('Model loaded')