Question

我想导入以下文件，该文件包含每周格式（仅星期四）的数据，并将其转换为每日文件，其值从星期四到下一个星期三（跳过星期六和星期日）填写。

https://www.aaii.com/files/surveys/sentiment.xls

我可以导入它：

df = pd.read_excel("C:\\Users\\Public\\Portfolio\\exports\\sentiment.xls", sheet_name = "SENTIMENT", skiprows=3, parse_dates=['Date'], date_format='%m-%d-%y')

这是结果：

但是那是我所能得到的。即使是最简单的重采样也会失败

TypeError: Only valid with DatetimeIndex, TimedeltaIndex or PeriodIndex, but got an instance of 'RangeIndex'

我尝试了df['Date'] = pd.to_datetime(df['Date'])和其他方法，但均未获得成功。

关于如何完成此工作的想法？

Answer 1

您可以尝试。.

df = pd.read_excel("sentiment.xls", sheet_name = "SENTIMENT", skiprows=3, parse_dates=['Date'], date_format='%m-%d-%y')

您的“日期”列具有NaN值，因此当您尝试转换为datetime时，它会失败。.

>>> df['Date']
0                       NaN
1       1987-06-26 00:00:00
2       1987-07-17 00:00:00
3       1987-07-24 00:00:00
4       1987-07-31 00:00:00

因此，您需要使用coerce来转换日期时间。.

>>> df['Date'] = pd.to_datetime(df['Date'], errors='coerce')

现在您的日期已处理..

>>> df['Date']
0             NaT
1      1987-06-26
2      1987-07-17
3      1987-07-24
4      1987-07-31
5      1987-08-07
6      1987-08-14
7      1987-08-21

现在将索引设置为“日期”列，然后按照注释中的说明重新采样：

>>> df.set_index('Date', inplace=True)
>>> df.head()
            Bullish  Neutral  Bearish  Total  Mov Avg  Spread   Average  +St. Dev.  - St. Dev.    High     Low   Close
Date
NaT             NaN      NaN      NaN    NaN      NaN     NaN       NaN        NaN         NaN     NaN     NaN     NaN
1987-06-26      NaN      NaN      NaN    NaN      NaN     NaN  0.382642   0.484295    0.280989     NaN     NaN     NaN
1987-07-17      NaN      NaN      NaN    NaN      NaN     NaN  0.382642   0.484295    0.280989  314.59  307.63  314.59
1987-07-24     0.36     0.50     0.14    1.0      NaN    0.22  0.382642   0.484295    0.280989  311.39  307.81  309.27
1987-07-31     0.26     0.48     0.26    1.0      NaN    0.00  0.382642   0.484295    0.280989  318.66  310.65  318.66

Answer 2

我认为这是正确的答案，可以转换为每日，去除非交易日和周六/周日。

import pandas as pd
from pandas.tseries.offsets import BDay
# read csv, use SENTIMENT sheet, drop the first three rows, parse dates to datetime, index on date
df = pd.read_excel("C:\\Users\\Public\\Portfolio\\exports\\sentiment.xls", sheet_name = "SENTIMENT", skiprows=3, parse_dates=['Date'], date_format='%m-%d-%y', index_col ='Date')
df = df[3:].asfreq('D', method='ffill') # skip 3 lines then expand to daily and fill forward
df = df[df.index.map(BDay().onOffset)] # strip non-trading weekdays
df = df[df.index.dayofweek < 5] # strip Saturdays and Sundays
print(df.head(250))

也许有一种更优雅的方法，但是可以完成工作。

如何将AII的每周数据重新采样为每日？

2 个答案: