# pylint: disable=consider-using-with
"""Collection of loggers that can be inserted into an optimizer as callback."""
import csv
from abc import ABC, abstractmethod
from datetime import datetime
from pathlib import Path
from typing import IO, Any, Dict, Iterable, List, Optional, Type, Union
import numpy as np
import yaml
from tensorwaves.interface import Estimator, Optimizer, ParameterValue
[docs]class Loadable(ABC):
[docs] @staticmethod
@abstractmethod
def load_latest_parameters(filename: Union[Path, str]) -> dict:
pass
[docs]class Callback(ABC):
"""Interface for callbacks such as `.CSVSummary`.
.. seealso:: :ref:`usage/step3:Custom callbacks`
"""
[docs] @abstractmethod
def on_optimize_start(self, logs: Optional[Dict[str, Any]] = None) -> None:
pass
[docs] @abstractmethod
def on_optimize_end(self, logs: Optional[Dict[str, Any]] = None) -> None:
pass
[docs] @abstractmethod
def on_iteration_end(
self, iteration: int, logs: Optional[Dict[str, Any]] = None
) -> None:
pass
[docs] @abstractmethod
def on_function_call_end(
self, function_call: int, logs: Optional[Dict[str, Any]] = None
) -> None:
pass
[docs]class CallbackList(Callback):
"""Class for combining `Callback` s.
Combine different `Callback` classes in to a chain as follows:
>>> from tensorwaves.optimizer.callbacks import (
... CallbackList, TFSummary, YAMLSummary
... )
>>> from tensorwaves.optimizer.minuit import Minuit2
>>> optimizer = Minuit2(
... callback=CallbackList([TFSummary(), YAMLSummary("fit_result.yml")])
... )
"""
def __init__(self, callbacks: Iterable[Callback]) -> None:
self.__callbacks: List[Callback] = []
for callback in callbacks:
self.__callbacks.append(callback)
@property
def callbacks(self) -> List[Callback]:
return list(self.__callbacks)
[docs] def __eq__(self, other: object) -> bool:
if isinstance(other, CallbackList):
return self.callbacks == other.callbacks
return False
[docs] def on_optimize_start(self, logs: Optional[Dict[str, Any]] = None) -> None:
for callback in self.__callbacks:
callback.on_optimize_start(logs)
[docs] def on_optimize_end(self, logs: Optional[Dict[str, Any]] = None) -> None:
for callback in self.__callbacks:
callback.on_optimize_end(logs)
[docs] def on_iteration_end(
self, iteration: int, logs: Optional[Dict[str, Any]] = None
) -> None:
for callback in self.__callbacks:
callback.on_iteration_end(iteration, logs)
[docs] def on_function_call_end(
self, function_call: int, logs: Optional[Dict[str, Any]] = None
) -> None:
for callback in self.__callbacks:
callback.on_function_call_end(function_call, logs)
[docs]class CSVSummary(Callback, Loadable):
"""Log fit parameters and the estimator value to a CSV file."""
def __init__(
self,
filename: Union[Path, str],
function_call_step_size: int = 1,
iteration_step_size: Optional[int] = None,
) -> None:
if iteration_step_size is None:
iteration_step_size = 0
if function_call_step_size <= 0 and iteration_step_size <= 0:
raise ValueError(
"either function call or interaction step size should > 0."
)
self.__function_call_step_size = function_call_step_size
self.__iteration_step_size = iteration_step_size
self.__latest_function_call: Optional[int] = None
self.__latest_iteration: Optional[int] = None
self.__writer: Optional[csv.DictWriter] = None
self.__filename = filename
self.__stream: Optional[IO] = None
def __del__(self) -> None:
_close_stream(self.__stream)
[docs] def on_optimize_start(self, logs: Optional[Dict[str, Any]] = None) -> None:
if logs is None:
raise ValueError(
f"{self.__class__.__name__} requires logs on optimize start"
" to determine header names"
)
if self.__function_call_step_size > 0:
self.__latest_function_call = 0
if self.__iteration_step_size > 0:
self.__latest_iteration = 0
_close_stream(self.__stream)
self.__stream = open(self.__filename, "w", newline="")
self.__writer = csv.DictWriter(
self.__stream,
fieldnames=list(self.__log_to_rowdict(logs)),
quoting=csv.QUOTE_NONNUMERIC,
)
self.__writer.writeheader()
[docs] def on_optimize_end(self, logs: Optional[Dict[str, Any]] = None) -> None:
if logs is not None:
self.__latest_function_call = None
self.__latest_iteration = None
self.__write(logs)
_close_stream(self.__stream)
[docs] def on_iteration_end(
self, iteration: int, logs: Optional[Dict[str, Any]] = None
) -> None:
self.__latest_iteration = iteration
if logs is None:
return
if (
self.__iteration_step_size is None
or self.__latest_iteration % self.__iteration_step_size != 0
):
return
self.__write(logs)
[docs] def on_function_call_end(
self, function_call: int, logs: Optional[Dict[str, Any]] = None
) -> None:
self.__latest_function_call = function_call
if logs is None:
return
if (
self.__function_call_step_size is None
or self.__latest_function_call % self.__function_call_step_size
!= 0
):
return
self.__write(logs)
def __write(self, logs: Dict[str, Any]) -> None:
if self.__writer is None:
raise ValueError(
f"{csv.DictWriter.__name__} has not been initialized"
)
row_dict = self.__log_to_rowdict(logs)
self.__writer.writerow(row_dict)
def __log_to_rowdict(self, logs: Dict[str, Any]) -> Dict[str, Any]:
output = {
"time": logs["time"],
"optimizer": logs["optimizer"],
"estimator_type": logs["estimator"]["type"],
"estimator_value": logs["estimator"]["value"],
**logs["parameters"],
}
if self.__latest_function_call is not None:
output = {
"function_call": self.__latest_function_call,
**output,
}
if self.__latest_iteration is not None:
output = {
"iteration": self.__latest_iteration,
**output,
}
return output
[docs] @staticmethod
def load_latest_parameters(filename: Union[Path, str]) -> dict:
def cast_non_numeric(value: str) -> Union[complex, float, str]:
# https://docs.python.org/3/library/csv.html#csv.QUOTE_NONNUMERIC
# does not work well for complex numbers
try:
return complex(value)
except ValueError:
try:
return float(value)
except ValueError:
return value
with open(filename) as stream:
reader = csv.DictReader(stream)
last_line = list(reader)[-1]
return {
name: cast_non_numeric(value) for name, value in last_line.items()
}
[docs]class TFSummary(Callback):
"""Log fit parameters and the estimator value to a `tf.summary`.
The logs can be viewed with `TensorBoard
<https://www.tensorflow.org/tensorboard>`_ via:
.. code-block:: shell
tensorboard --logdir logs
"""
def __init__(
self,
logdir: str = "logs",
step_size: int = 10,
subdir: Optional[str] = None,
) -> None:
self.__logdir = logdir
self.__subdir = subdir
self.__step_size = step_size
self.__stream: Optional[Any] = None
[docs] def on_optimize_start(self, logs: Optional[Dict[str, Any]] = None) -> None:
# pylint: disable=import-outside-toplevel, no-member
import tensorflow as tf
output_dir = (
self.__logdir + "/" + datetime.now().strftime("%Y%m%d-%H%M%S")
)
if self.__subdir is not None:
output_dir += "/" + self.__subdir
self.__stream = tf.summary.create_file_writer(output_dir)
self.__stream.set_as_default() # type: ignore[attr-defined]
[docs] def on_optimize_end(self, logs: Optional[Dict[str, Any]] = None) -> None:
if self.__stream:
self.__stream.close()
[docs] def on_iteration_end(
self, iteration: int, logs: Optional[Dict[str, Any]] = None
) -> None:
pass
[docs] def on_function_call_end(
self, function_call: int, logs: Optional[Dict[str, Any]] = None
) -> None:
# pylint: disable=import-outside-toplevel, no-member
import tensorflow as tf
if logs is None:
return
if function_call % self.__step_size != 0:
return
parameters = logs["parameters"]
for par_name, value in parameters.items():
tf.summary.scalar(par_name, value, step=function_call)
estimator_value = logs.get("estimator", {}).get("value", None)
if estimator_value is not None:
tf.summary.scalar("estimator", estimator_value, step=function_call)
if self.__stream is not None:
self.__stream.flush()
[docs]class YAMLSummary(Callback, Loadable):
"""Log fit parameters and the estimator value to a `tf.summary`.
The logs can be viewed with `TensorBoard
<https://www.tensorflow.org/tensorboard>`_ via:
.. code-block:: shell
tensorboard --logdir logs
"""
def __init__(
self, filename: Union[Path, str], step_size: int = 10
) -> None:
self.__step_size = step_size
self.__filename = filename
self.__stream: Optional[IO] = None
def __del__(self) -> None:
_close_stream(self.__stream)
[docs] def on_optimize_start(self, logs: Optional[Dict[str, Any]] = None) -> None:
_close_stream(self.__stream)
self.__stream = open(self.__filename, "w")
[docs] def on_optimize_end(self, logs: Optional[Dict[str, Any]] = None) -> None:
if logs is None:
return
self.__dump_to_yaml(logs)
_close_stream(self.__stream)
[docs] def on_iteration_end(
self, iteration: int, logs: Optional[Dict[str, Any]] = None
) -> None:
pass
[docs] def on_function_call_end(
self, function_call: int, logs: Optional[Dict[str, Any]] = None
) -> None:
if logs is None:
return
if function_call % self.__step_size != 0:
return
self.__dump_to_yaml(logs)
def __dump_to_yaml(self, logs: Dict[str, Any]) -> None:
_empty_file(self.__stream)
cast_logs = dict(logs)
cast_logs["parameters"] = {
p: _cast_value(v) for p, v in logs["parameters"].items()
}
yaml.dump(
cast_logs,
self.__stream,
sort_keys=False,
Dumper=_IncreasedIndent,
default_flow_style=False,
)
[docs] @staticmethod
def load_latest_parameters(filename: Union[Path, str]) -> dict:
with open(filename) as stream:
fit_stats = yaml.load(stream, Loader=yaml.Loader)
return fit_stats["parameters"]
def _cast_value(value: Any) -> ParameterValue:
# cspell:ignore iscomplex
if np.iscomplex(value) or isinstance(value, complex):
return complex(value)
return float(value)
class _IncreasedIndent(yaml.Dumper):
# pylint: disable=too-many-ancestors
def increase_indent(
self, flow: bool = False, indentless: bool = False
) -> None:
return super().increase_indent(flow, False)
def _close_stream(stream: Optional[IO]) -> None:
if stream is not None:
stream.close()
def _empty_file(stream: Optional[IO]) -> None:
if stream is None:
return
stream.seek(0)
stream.truncate()
def _create_log( # pyright: reportUnusedFunction=false
optimizer: Type[Optimizer],
estimator_value: float,
estimator_type: Type[Estimator],
parameters: Dict[str, Any],
function_call: int,
) -> Dict[str, Any]:
return {
"time": datetime.now(),
"optimizer": optimizer.__name__,
"estimator": {
"type": estimator_type.__name__,
"value": float(estimator_value),
},
"function_call": function_call,
"parameters": parameters,
}
