I'm using daily data but trade entries are duplicating with different exit times also

[BlackArbsCEO]

I'm using the example machine learning script from here: original backtesting docs ml example

where I use super().next() to modify the original next. However, using daily data, I get duplicate trades for the same entry timestamp, sometimes with different sizes and exit times.

Some example code:

    class MLTrainOnceStrategy(Strategy):

        train_year_end = TRAIN_YEAR_END # 2012
        n_train = N_TRAIN  # 1000
        retrain = N_RETRAIN  # 100
        max_position_size = 0.05
        min_position_size = 0.01
        current_date = None
        current_dates = [np.nan]

        def init(self):
            # Init our model
            self.clf = create_model()

            # Train the classifier in advance on the first N_TRAIN examples
            df = self.data.df.loc[: str(self.train_year_end)]

            X, y = get_clean_Xy(df, get_feature_columns(df, Y_NAME), Y_NAME)
            self.clf.fit(X, y)

            # Plot y for inspection
            self.I(get_y, self.data.df, name=Y_NAME)

            # Prepare empty, all-NaN forecast indicator
            self.forecasts = self.I(
                lambda: np.repeat(np.nan, len(self.data)), name="forecast"
            )

        def next(self):

            if int(self.data.index.year[-1]) < int(self.train_year_end):
                return
            # Proceed only with out-of-sample data. Prepare some variables
            high, low, close = self.data.High, self.data.Low, self.data.Close
            current_time = self.data.index[-1]

            if current_time.date() == self.current_dates[-1]:
                return

            # Forecast the next movement
            X = get_X(
                self.data.df.iloc[-1:],
                get_feature_columns(self.data.df.iloc[-1:], Y_NAME),
            )
            forecast = self.clf.predict(X)[0]
            if forecast == 1:
                forecast_proba = self.clf.predict_proba(X)[0][-1]
            elif forecast == -1:
                forecast_proba = self.clf.predict_proba(X)[0][-1]  

            # Update the plotted "forecast" indicator
            self.forecasts[-1] = forecast_proba

            if forecast == 1 and not self.position.is_long:
                allocation = max(
                    self.min_position_size, abs(self.max_position_size * forecast_proba)
                )
                self.buy(size=allocation)
            elif forecast == -1 and not self.position.is_short:
                allocation = max(
                    self.min_position_size, abs(self.max_position_size * forecast_proba)
                )
                self.sell(size=allocation)
            self.current_date = current_time.date()
            self.current_dates.append(self.current_date)
            # print(f"REGULAR NEXT: {self.current_dates=}")

    class MLWalkForwardStrategy(MLTrainOnceStrategy):
        def next(self):
            if int(self.data.index.year[-1]) < int(self.train_year_end):
                return

            current_time = self.data.index[-1]

            if current_time.date() == self.current_dates[-1]:
                return

            if len(self.data) % int(self.retrain):
                return super().next()

            df = self.data.df.iloc[-int(self.n_train) :]
            if not df[df.duplicated()].empty:
                raise ValueError("wtf data df has duplicate data")

            X, y = get_clean_Xy(df, get_feature_columns(df, Y_NAME), Y_NAME)
            self.clf.fit(X, y)
            self.current_date = current_time.date()
            self.current_dates.append(self.current_date)
            super().next()

Again with daily data, the trades are duplicated:

Size	EntryBar	ExitBar	EntryPrice	ExitPrice	PnL	ReturnPct	EntryTime	ExitTime	Tag	Duration
0	540	3205	3243	8.538252	9.561705	552.664532	0.119867	2012-01-04 00:00:00-05:00	2012-02-29 00:00:00-05:00	None	56 days 00:00:00
1	549	3205	3244	8.538252	9.542361	551.255804	0.117601	2012-01-04 00:00:00-05:00	2012-03-01 00:00:00-05:00	None	57 days 00:00:00
2	-9	3244	3247	9.540453	9.400515	1.259439	0.014668	2012-03-01 00:00:00-05:00	2012-03-06 00:00:00-05:00	None	5 days 00:00:00
3	515	3247	3250	9.402395	9.555253	78.721937	0.016257	2012-03-06 00:00:00-05:00	2012-03-09 00:00:00-05:00	None	3 days 00:00:00
4	78	3247	3251	9.402395	9.600389	15.443531	0.021058	2012-03-06 00:00:00-05:00	2012-03-12 00:00:00-04:00	None	5 days 23:00:00
5	-410	3251	3256	9.598469	10.167896	-233.465038	-0.059325	2012-03-12 00:00:00-04:00	2012-03-19 00:00:00-04:00	None	7 days 00:00:00
6	213	3256	3271	10.169930	9.883297	-61.052686	-0.028184	2012-03-19 00:00:00-04:00	2012-04-10 00:00:00-04:00	None	22 days 00:00:00
7	221	3256	3272	10.169930	9.818618	-77.639856	-0.034544	2012-03-19 00:00:00-04:00	2012-04-11 00:00:00-04:00	None	23 days 00:00:00
8	-62	3272	3293	9.816654	9.773341	2.685418	0.004412	2012-04-11 00:00:00-04:00	2012-05-10 00:00:00-04:00	None	29 days 00:00:00