for testing and deploying your application
for finding and fixing issues
for empowering human code reviews
import pandas as pd
from collections.abc import Sequence
class ResultsManager:
def __init__(self):
self._results_l = []
def add(self, result, params) -> None:
score_d = {"score": result.score}
results_dict = {**score_d, **result.metrics}
self._results_l.append({**results_dict, **params})
@property
def dataframe(self) -> pd.DataFrame:
return pd.DataFrame(self._results_l)
def best(self):
return max(self._results, key=lambda r: r.score, default=None)
SimonBlanke /
Gradient-Free-Optimizers
