Coverage for mlos_bench/mlos_bench/storage/base_tunable_config_trial_group_data.py: 97%
37 statements
« prev ^ index » next coverage.py v7.5.1, created at 2024-05-06 00:35 +0000
« prev ^ index » next coverage.py v7.5.1, created at 2024-05-06 00:35 +0000
1#
2# Copyright (c) Microsoft Corporation.
3# Licensed under the MIT License.
4#
5"""
6Base interface for accessing the stored benchmark config trial group data.
7"""
9from abc import ABCMeta, abstractmethod
10from typing import Any, Dict, Optional, TYPE_CHECKING
12import pandas
14from mlos_bench.storage.base_tunable_config_data import TunableConfigData
16if TYPE_CHECKING:
17 from mlos_bench.storage.base_trial_data import TrialData
20class TunableConfigTrialGroupData(metaclass=ABCMeta):
21 """
22 Base interface for accessing the stored experiment benchmark tunable config
23 trial group data.
25 A (tunable) config is used to define an instance of values for a set of tunable
26 parameters for a given experiment and can be used by one or more trial instances
27 (e.g., for repeats), which we call a (tunable) config trial group.
28 """
30 def __init__(self, *,
31 experiment_id: str,
32 tunable_config_id: int,
33 tunable_config_trial_group_id: Optional[int] = None):
34 self._experiment_id = experiment_id
35 self._tunable_config_id = tunable_config_id
36 # can be lazily initialized as necessary:
37 self._tunable_config_trial_group_id: Optional[int] = tunable_config_trial_group_id
39 @property
40 def experiment_id(self) -> str:
41 """
42 ID of the experiment.
43 """
44 return self._experiment_id
46 @property
47 def tunable_config_id(self) -> int:
48 """
49 ID of the config.
50 """
51 return self._tunable_config_id
53 @abstractmethod
54 def _get_tunable_config_trial_group_id(self) -> int:
55 """
56 Retrieve the trial's config_trial_group_id from the storage.
57 """
58 raise NotImplementedError("subclass must implement")
60 @property
61 def tunable_config_trial_group_id(self) -> int:
62 """
63 The unique ID (within this experiment) of the (tunable) config trial group.
65 This is a unique identifier for all trials in this experiment using the given
66 config_id, and typically defined as the the minimum trial_id for the given
67 config_id.
68 """
69 if self._tunable_config_trial_group_id is None:
70 self._tunable_config_trial_group_id = self._get_tunable_config_trial_group_id()
71 assert self._tunable_config_trial_group_id is not None
72 return self._tunable_config_trial_group_id
74 def __repr__(self) -> str:
75 return f"TunableConfigTrialGroup :: {self._experiment_id} cid:{self.tunable_config_id}"
77 def __eq__(self, other: Any) -> bool:
78 if not isinstance(other, self.__class__):
79 return False
80 return self._tunable_config_id == other._tunable_config_id and self._experiment_id == other._experiment_id
82 @property
83 @abstractmethod
84 def tunable_config(self) -> TunableConfigData:
85 """
86 Retrieve the (tunable) config data for this (tunable) config trial group from the storage.
88 Returns
89 -------
90 TunableConfigData
91 """
93 @property
94 @abstractmethod
95 def trials(self) -> Dict[int, "TrialData"]:
96 """
97 Retrieve the trials' data for this (tunable) config trial group from the storage.
99 Returns
100 -------
101 trials : Dict[int, TrialData]
102 A dictionary of the trials' data, keyed by trial id.
103 """
105 @property
106 @abstractmethod
107 def results_df(self) -> pandas.DataFrame:
108 """
109 Retrieve all results for this (tunable) config trial group as a single DataFrame.
111 Returns
112 -------
113 results : pandas.DataFrame
114 A DataFrame with configurations and results from all trials of the experiment.
115 Has columns [trial_id, config_id, ts_start, ts_end, status]
116 followed by tunable config parameters (prefixed with "config.") and
117 trial results (prefixed with "result."). The latter can be NULLs if the
118 trial was not successful.
120 See Also
121 --------
122 ExperimentData.results
123 """