Logging

potpourri uses loguru for structured logging. Following the library best-practice, all output is suppressed by default so that the library never pollutes your application's log stream.

Enabling logging

from loguru import logger
logger.enable("potpourri")

Call this once at the start of your script. All subsequent potpourri calls will emit log records to loguru's default sink (stderr with colours).

Log levels

Level	What is logged
`DEBUG`	Low-level model internals: solver options, variable fix/unfix operations, result mapping to `net.res_*`
`INFO`	Normal milestones: model creation timestamp, solve started (with solver name), optimal solution found, NEOS job submitted
`WARNING`	Non-fatal issues that may affect results: generator voltage limits overriding bus limits, missing `p_inst_mw` attribute (fallback used), non-optimal solver termination, unknown model component in `change_vals` / `fix_vars`
`ERROR`	Caught exceptions: solver errors (mindtpy `ValueError`), termination-condition check failures, `fix_vars` / `unfix_vars` attribute errors

Filtering by level

Show only INFO and above (suppress DEBUG):

import sys
from loguru import logger

logger.remove()                          # remove the default sink
logger.add(sys.stderr, level="INFO")     # add it back at INFO
logger.enable("potpourri")

Show only WARNING and above (quiet normal operation):

logger.remove()
logger.add(sys.stderr, level="WARNING")
logger.enable("potpourri")

Show everything including DEBUG:

logger.remove()
logger.add(sys.stderr, level="DEBUG")
logger.enable("potpourri")

Writing logs to a file

logger.add("potpourri.log", level="DEBUG", rotation="10 MB")
logger.enable("potpourri")

loguru rotates the file automatically once it reaches 10 MB. See the loguru docs for retention and compression options.

Disabling logging again

logger.disable("potpourri")