Skip to content
Snippets Groups Projects
defaults.rst 1.74 KiB

Defaults

In this section, we explain which parameters are set by MLAir during the :py:`ExperimentSetup` if not specified by the user. This is important information for example if a new :ref:`Custom Data Handler` is implemented.

parameter default comment
batch_path    
batch_size    
bootstrap_path    
competitor_path    
competitors    
create_new_bootstraps    
create_new_model    
data_handler    
data_origin    
data_path    
debug
MLAir checks if it is running in debug mode and stores this
dimensions    
end    
epochs    
evaluate_bootstraps    
experiment_name    
experiment_path    
extreme_values    
extremes_on_right_tail_only    
forecast_path    
fraction_of_training    
hostname    
hpc_hosts    
interpolation_limit    
interpolation_method    
logging_path    
login_nodes    
model_class    
model_path    
neighbors    
number_of_bootstraps    
overwrite_local_data    
permute_data    
plot_list    
plot_path    
start    
stations    
statistics_per_var    
target_dim    
target_var    
test_start    
test_end    
test_min_length    
time_dim    
train_model    
train_end    
train_min_length    
train_start    
transformation :py:`{}` implement all further transformation functionality inside your custom data handler
use_all_stations_on_all_data_sets    
use_multiprocessing :py:`True` is set to False if MLAir is running in debug mode
upsampling    
val_end    
val_min_length    
val_start    
variables    
window_history_size    
window_lead_time