提取符合条件的数据，代码优化，提升计算效率

这段代码可以怎么优化呀
运行计算效率低得离谱，我也找不出问题在哪里


    for i in range(0, (endtime - starttime).seconds + 1):
        tradingtime = starttime + timedelta(seconds=i)
        if str(tradingtime.time()) not in time1:
            print(tradingtime)
            try:
                underlying_510300 = price_data['510300'].get_group(tradingtime).tail(1)['pre_close'].values
                underlying_000300 = price_data['000300'].get_group(tradingtime).tail(1)['pre_close'].values
            except Exception as e:
                underlying_510300 = \
                price.get_group('510300')[price.get_group('510300')['inserttime'].dt.time <= tradingtime.time()].tail(
                    1)['pre_close'].values
                underlying_000300 = \
                price.get_group('000300')[price.get_group('000300')['inserttime'].dt.time <= tradingtime.time()].tail(
                    1)['pre_close'].values
            MKT_data = {}
            for code in code_group:
                try:
                    option_data = price_data[code].get_group(tradingtime).tail(1)
                except Exception as e:
                    option_data = price.get_group(code)[
                        price.get_group(code)['inserttime'].dt.time <= tradingtime.time()].tail(1)
                if INFO_group[code]['underlying'] == '510300':
                    option_data['underlying_price'] = underlying_510300
                else:
                    option_data['underlying_price'] = underlying_000300
                MKT_data[code] = option_data

参考gpt以及个人思路优化过程：
1.在每个循环中，都会根据时间戳来查询数据，可以将查询的结果缓存起来，以避免重复查询。

2.在异常处理中使用了两次查询，可以将两次查询合并成一次，以避免重复代码。

3.可以尝试使用并行化的方式来加速数据处理过程。

代码如下：

from concurrent.futures import ThreadPoolExecutor

def get_price_data(code, tradingtime, price_data, price):
    try:
        option_data = price_data[code].get_group(tradingtime).tail(1)
    except Exception as e:
        option_data = price.get_group(code)[
            price.get_group(code)['inserttime'].dt.time <= tradingtime.time()].tail(1)
    return option_data

def optimize_code(starttime, endtime, time1, price_data, price, code_group):
    underlying_510300 = None
    underlying_000300 = None
    MKT_data = {}

    pool = ThreadPoolExecutor(max_workers=8)
    underlying_future_510300 = pool.submit(price_data['510300'].get_group, starttime)
    underlying_future_000300 = pool.submit(price_data['000300'].get_group, starttime)
    for i in range(0, (endtime - starttime).seconds + 1):
        tradingtime = starttime + timedelta(seconds=i)
        if str(tradingtime.time()) not in time1:
            if underlying_510300 is None:
                underlying_510300 = underlying_future_510300.result().tail(1)['pre_close'].values
                underlying_000300 = underlying_future_000300.result().tail(1)['pre_close'].values
            try:
                option_data = price_data[code].get_group(tradingtime).tail(1)
            except Exception as e:
                option_data = price.get_group(code)[
                    price.get_group(code)['inserttime'].dt.time <= tradingtime.time()].tail(1)

            if code not in MKT_data:
                MKT_data[code] = {}
            MKT_data[code]['option_data'] = option_data
            if INFO_group[code]['underlying'] == '510300':
                if underlying_510300 is not None:
                    MKT_data[code]['underlying_price'] = underlying_510300
            else:
                if underlying_000300 is not None:
                    MKT_data[code]['underlying_price'] = underlying_000300

    return MKT_data