MCPcopy Create free account
hub / github.com/antmachineintelligence/mtgbmcode / record_evaluation

Function record_evaluation

python-package/lightgbmmt/callback.py:78–106  ·  view source on GitHub ↗

Create a callback that records the evaluation history into ``eval_result``. Parameters ---------- eval_result : dict A dictionary to store the evaluation results. Returns ------- callback : function The callback that records the evaluation history into the pa

(eval_result)

Source from the content-addressed store, hash-verified

76
77
78def record_evaluation(eval_result):
79 """Create a callback that records the evaluation history into ``eval_result``.
80
81 Parameters
82 ----------
83 eval_result : dict
84 A dictionary to store the evaluation results.
85
86 Returns
87 -------
88 callback : function
89 The callback that records the evaluation history into the passed dictionary.
90 """
91 if not isinstance(eval_result, dict):
92 raise TypeError('eval_result should be a dictionary')
93 eval_result.clear()
94
95 def _init(env):
96 for data_name, eval_name, _, _ in env.evaluation_result_list:
97 eval_result.setdefault(data_name, collections.OrderedDict())
98 eval_result[data_name].setdefault(eval_name, [])
99
100 def _callback(env):
101 if not eval_result:
102 _init(env)
103 for data_name, eval_name, result, _ in env.evaluation_result_list:
104 eval_result[data_name][eval_name].append(result)
105 _callback.order = 20
106 return _callback
107
108
109def reset_parameter(**kwargs):

Callers

nothing calls this directly

Calls 1

clearMethod · 0.45

Tested by

no test coverage detected