logml.log

Functions

get_default_logger([log_file, log_format, ...])	Returns default logger.
init_logger(global_params)	Initialize logger with global params:
log_header(log_f, message, *args)	Log message wrapped around two long dashed lines.
prepare_logger([logger, log_file, ...])	Basic logger configuration: log format, handles setup.

logml.log.log_header(log_f, message: str, *args)

Log message wrapped around two long dashed lines.

logml.log.get_default_logger(log_file: Optional[str] = None, log_format: Optional[str] = None, level=None, extra: Optional[dict] = None, use_console=True)

Returns default logger. If not yet initialized, prepare_logger is used.

logml.log.prepare_logger(logger=None, log_file: Optional[str] = None, log_format: Optional[str] = None, level=None, extra: Optional[dict] = None, use_console=True)

Basic logger configuration: log format, handles setup.

Parameters

• use_console – When True, then logging is performed to the console.

• logger – Pre-configured logger object. If not provided, will be created from scratch as logging.getLogger(‘logml’).

• log_file – Log file name. If not provided, logging is performed only to console.

• log_format (str) – Log format string. See standard logging formatting rules.

• level – Logging level

• extra –

  extra data to log, should be controlled by log_format. Example:

  logger = prepare_logger(
      ...
      log_format="%(process_id)s %(model_name)s --- %(message)s",
      extra=dict(process_id=123, model_name='lasso')
  )
  logger.info('m2')
  

  results in

  123 lasso --- m2
  

Returns

logger object (possibly wrapped with adapter, if extra is provided.)

logml.log.init_logger(global_params: dict)

Initialize logger with global params:

global_params used:

job_id: if set, will be included into logger format.

log_file: if specified, will be used as log file.

If only a filename specified, will be put to destintion folder If omited, default filename is ${destination_folder}/logs/logml.log

Returns

tuple (logger, log_file_name)