TR指标计算异常：首日TR2/TR3未显示NaN及数据范围疑问

问题背景 #

数据起始日期为2024-03-27，该日期无前置交易日数据，按照TR指标逻辑，TR2和TR3应返回NaN，但实际输出了有效数值。要求严格基于当前展示的日期区间完成指标计算。

初步排查 #

• 验证shift(-1)函数逻辑，计算结果符合预期，排除函数使用错误
• 怀疑AlphaVantage API返回的数据超出了设定的4天范围，导致起始日期仍能获取到前置数据，进而使TR2、TR3计算出有效值

修改后的calculate_tr函数 #

def calculate_tr(df):
    high_shifted, close_shifted, low_shifted = (
        df["High"].shift(),
        df["Close"].shift(),
        df["Low"].shift(),
    )

    df["tr1"] = df["High"] - df["Low"]
    df["tr2"] = (high_shifted - close_shifted).abs()
    df["tr3"] = (low_shifted - close_shifted).abs()

输出结果 #

原脚本 #

import requests
import pandas as pd
from datetime import datetime, timedelta
import time


# 计算每日历史数据的起止日期
end_date_daily = datetime.now().strftime('%Y-%m-%d')
start_date_daily = (datetime.now() - timedelta(days=4)).strftime('%Y-%m-%d')

# 构建每日历史数据的API请求URL
api_url_daily = f'https://www.alphavantage.co/query?function=TIME_SERIES_DAILY&symbol={symbol}&apikey={API_KEY}&datatype=json&start_date={start_date_daily}&end_date={end_date_daily}'

# 获取并处理每日历史价格数据
response_daily = requests.get(api_url_daily)

if response_daily.status_code == 200:
    historical_data_daily = response_daily.json()['Time Series (Daily)']

    # 将每日历史数据转换为DataFrame
    df_daily = pd.DataFrame(historical_data_daily).T
    df_daily.index = pd.to_datetime(df_daily.index)

    # 反转DataFrame，使最早数据位于顶部
    df_daily = df_daily.sort_index(ascending=False)

    # 转换列类型为浮点型
    df_daily['High'] = df_daily['2. high'].astype(float)
    df_daily['Low'] = df_daily['3. low'].astype(float)
    df_daily['Close'] = df_daily['4. close'].astype(float)


    # 计算tr1、tr2、tr3的函数
    def calculate_tr(df):
        df['tr1'] = df['High'] - df['Low']
        df['tr2'] = abs(df['High'] - df['Close'].shift(-1))
        df['tr3'] = abs(df['Low'] - df['Close'].shift(-1))

    # 计算tr1、tr2、tr3
    calculate_tr(df_daily)

    # 输出计算后的DataFrame
    print("\n计算tr1、tr2、tr3后的DataFrame:")
    print("日期       开盘价  最高价  最低价  收盘价  TR1  TR2  TR3")
    for date in reversed(df_daily.head(4).index):
        row = df_daily.loc[date]
        tr1_rounded = round(row['tr1'], 4)
        tr2_rounded = round(row['tr2'], 4)
        tr3_rounded = round(row['tr3'], 4)
        print(f"{date.strftime('%Y-%m-%d')} {row['1. open']} {row['2. high']} {row['3. low']} {row['4. close']} {round(tr1_rounded, 3)} {round(tr2_rounded, 3)} {round(tr3_rounded, 3)}")

内容的提问来源于stack exchange，提问作者bob mcgee