操纵 Dataframe 从最近的变化午夜时间戳开始

Question

我的目标：

我有一个每天随机生成的数据集，导致第一行在随机时间开始。 我想让这个数据集从最近的午夜日期开始。 例如，如果第一行的日期是2022-05-09 15:00:00 ，我将不得不对数据进行切片以使其从最近的午夜开始，在这种情况下： 2022-05-10 00:00:00

这是数据集的样子：

我试过的：

我想找到我想要的时间戳第一次出现的索引，然后应用iloc来创建想要的数据集。

match_timestamp = "00:00:00"
[df[df.index.strftime("%H:%M:%S") == match_timestamp].first_valid_index()]

结果：[时间戳（'2022-05-10 00:00:00'）]

但是，这只会导致提取首次出现的时间戳，而且我无法将iloc应用于行值。 截至目前，我被困住了，想不出一个更优雅的解决方案，我确信它存在。

如果您能为此推荐一个更好的方法，我将不胜感激。 先感谢您！

这是提取 df 的完整代码：

pip 安装ccxt

import pandas as pd
import ccxt

exchange = ccxt.okx({'options': {'defaultType': 'futures', 'enableRateLimit': True}})
markets = exchange.load_markets()

url = 'https://www.okex.com'
tickers = pd.DataFrame((requests.get(url+'/api/v5/market/tickers?instType=FUTURES').json())['data'])
tickers = tickers.drop('instType', axis=1)
futures_tickers = list(tickers['instId'])

symbol = 'LINK-USD-220930'
candlestick_chart= exchange.fetch_ohlcv(symbol, '1h', limit=500)
candlestick_df = pd.DataFrame(candlestick_chart)
candlestick_df.columns = ['date', 'open', 'high', 'low', 'close', 'volume']

candlestick_df['date'] = pd.to_datetime(candlestick_df['date'], unit='ms')
candlestick_df['date'] = candlestick_df['date'] + pd.Timedelta(hours=8)

df  = candlestick_df
df

字典格式：（如建议的那样）

{'open': {Timestamp('2022-05-09 15:00:00'): 9.742, Timestamp('2022-05-09 16:00:00'): 9.731, Timestamp('2022-05-09 17:00:00'): 9.743, Timestamp('2022-05-09 18:00:00'): 9.684, Timestamp('2022-05-09 19:00:00'): 9.206, Timestamp('2022-05-09 20:00:00'): 9.43, Timestamp('2022-05-09 21:00:00'): 9.316, Timestamp('2022-05-09 22:00:00'): 9.403, Timestamp('2022-05-09 23:00:00'): 9.215, Timestamp('2022-05-10 00:00:00'): 9.141}, 'high': {Timestamp('2022-05-09 15:00:00'): 9.835, Timestamp('2022-05-09 16:00:00'): 9.75, Timestamp('2022-05-09 17:00:00'): 9.788, Timestamp('2022-05-09 18:00:00'): 9.697, Timestamp('2022-05-09 19:00:00'): 9.465, Timestamp('2022-05-09 20:00:00'): 9.469, Timestamp('2022-05-09 21:00:00'): 9.515, Timestamp('2022-05-09 22:00:00'): 9.413, Timestamp('2022-05-09 23:00:00'): 9.308, Timestamp('2022-05-10 00:00:00'): 9.223}, 'low': {Timestamp('2022-05-09 15:00:00'): 9.699, Timestamp('2022-05-09 16:00:00'): 9.596, Timestamp('2022-05-09 17:00:00'): 9.674, Timestamp('2022-05-09 18:00:00'): 8.739, Timestamp('2022-05-09 19:00:00'): 9.11, Timestamp('2022-05-09 20:00:00'): 9.3, Timestamp('2022-05-09 21:00:00'): 9.208, Timestamp('2022-05-09 22:00:00'): 9.174, Timestamp('2022-05-09 23:00:00'): 9.035, Timestamp('2022-05-10 00:00:00'): 8.724}, 'close': {Timestamp('2022-05-09 15:00:00'): 9.725, Timestamp('2022-05-09 16:00:00'): 9.745, Timestamp('2022-05-09 17:00:00'): 9.682, Timestamp('2022-05-09 18:00:00'): 9.18, Timestamp('2022-05-09 19:00:00'): 9.426, Timestamp('2022-05-09 20:00:00'): 9.32, Timestamp('2022-05-09 21:00:00'): 9.397, Timestamp('2022-05-09 22:00:00'): 9.229, Timestamp('2022-05-09 23:00:00'): 9.152, Timestamp('2022-05-10 00:00:00'): 8.82}, 'volume': {Timestamp('2022-05-09 15:00:00'): 3663.0, Timestamp('2022-05-09 16:00:00'): 6603.0, Timestamp('2022-05-09 17:00:00'): 2855.0, Timestamp('2022-05-09 18:00:00'): 20084.0, Timestamp('2022-05-09 19:00:00'): 8972.0, Timestamp('2022-05-09 20:00:00'): 5551.0, Timestamp('2022-05-09 21:00:00'): 8218.0, Timestamp('2022-05-09 22:00:00'): 7651.0, Timestamp('2022-05-09 23:00:00'): 6935.0, Timestamp('2022-05-10 00:00:00'): 10409.0}}

Answer 1

我对pandaNewstarter的简约方法，您只需将其应用到您的candlestick_df即可：

import pandas as pd
import datetime
df = pd.read_csv("data.csv")
df.dtypes
# convert date column to dtype timestamp
df.date = pd.to_datetime(df.date) 

# get min value from date colum
min_date = df.date.min()

# from min get next day midnight timestamp value
NextDay_Date = (min_date + datetime.timedelta(days=1)).replace(hour=0, minute=0, second=0, microsecond=0)
NextDay_Date

# create new DateFrame by slicing original
df2 = df[df.date >= NextDay_Date].copy()

出去：

print(NextDay_Date)
2022-05-10 00:00:00
print(df2)
                   date   open   high    low  close   volume
9   2022-05-10 00:00:00  9.141  9.223  8.724  8.820  10409.0
10  2022-05-10 01:00:00  8.755  8.979  8.558  8.832  11522.0
11  2022-05-10 02:00:00  8.815  8.880  8.304  8.593  20969.0
12  2022-05-10 03:00:00  8.618  8.720  8.370  8.610  15794.0
13  2022-05-10 04:00:00  8.610  8.929  8.610  8.736   9410.0
..                  ...    ...    ...    ...    ...      ...