Optuna：如何在研究中建议 function 和操作员？

Question

给定一个具有 4 列x1 - x3的数据框，我想编写一个 function 来近似最后一列true 。

我希望能够对列和不同的运算符尝试不同的功能。

我如何用 optuna 做到这一点？

可重现的尝试：

class Objective:

    def __init__(self):

    def __call__(self, trial):

# 1. Suggestion operands

        p1 = trial.suggest_categorical("op", [*,/,-,+])
        p2 = trial.suggest_categorical("op", [*,/,-,+])

# 2. Suggest functions

        f = trial.suggest_categorical("f", [exp(),log(),sin(),None,**2])

# 3. try all placements of the functions and operators to make a function to approximate a prediction
        predictions = f data['x1'] p1 data['x2'] p2 data['x3']

        score = accuracy(predictions, data["true"])
        return score

import optuna
objective = Objective()

# Setting SEED 
from optuna.samplers import TPESampler
sampler = TPESampler(seed=10)

study = optuna.create_study(
    pruner=optuna.pruners.MedianPruner(n_warmup_steps=10), direction="maximize",
    sampler=sampler
)
study.optimize(objective, n_trials=1500)