我有此代码，需要计算：[diff = max -min]。我在I/m试图减去的代码的最后一行中遇到了错误。我已经使用了pd.concat函数，应该使用合并函数吗？
import numpy as np
import pandas as pd
ts = ['25/02/2023 0:00', '25/02/2023 0:01', '25/02/2023 0:02',
'25/02/2023 0:03', '25/02/2023 0:04', '25/02/2023 0:05',
'25/02/2023 0:06', '25/02/2023 0:07', '25/02/2023 0:08',
'25/02/2023 0:09', '25/02/2023 0:10', '25/02/2023 0:11',
'25/02/2023 0:12', '25/02/2023 0:13', '25/02/2023 0:14',
'25/02/2023 0:15', '25/02/2023 0:16', '25/02/2023 0:17',
'25/02/2023 0:18', '25/02/2023 0:19', '25/02/2023 0:20',
'25/02/2023 0:21', '25/02/2023 0:22', '25/02/2023 0:23',
'25/02/2023 0:24', '25/02/2023 0:25', '25/02/2023 0:26',
'25/02/2023 0:27', '25/02/2023 0:28', '25/02/2023 0:29',
'25/02/2023 0:30', '25/02/2023 0:31', '25/02/2023 0:32',
'25/02/2023 0:33', '25/02/2023 0:34', '25/02/2023 0:35',
'25/02/2023 0:36', '25/02/2023 0:37', '25/02/2023 0:38',
'25/02/2023 0:39', '25/02/2023 0:40', '25/02/2023 0:41',
'25/02/2023 0:42', '25/02/2023 0:43', '25/02/2023 0:44',
'25/02/2023 0:45', '25/02/2023 0:46', '25/02/2023 0:47',
'25/02/2023 0:48', '25/02/2023 0:49', '25/02/2023 0:50',
'25/02/2023 0:51', '25/02/2023 0:52', '25/02/2023 0:53',
'25/02/2023 0:54', '25/02/2023 0:55', '25/02/2023 0:56',
'25/02/2023 0:57', '25/02/2023 0:58', '25/02/2023 0:59',
'25/02/2023 1:00']
temp = ['0', '21', '20', '30', '40', '50', '6', '7', '8', '9',
'10', '11', '12', '13', '14', '15', '16', '17', '18',
'19', '20', '21', '22', '23', '24', '25', '26', '27',
'28', '29', '68', '31', '32', '33', '34', '35', '36',
'37', '38', '39', '40', '41', '42', '43', '44', '45',
'46', '47', '48', '49', '50', '51', '52', '53', '54',
'55', '56', '57', '58', '59', '60', '61', '62']
df = pd.DataFrame(list(zip(ts, temp)),
columns = ['ts', 'temp'])
df['ts'] = pd.to_datetime(df['ts'])
df1 = df.set_index('ts')
print(df1)
df2 = df1.rolling(1, step=1).agg(['min'])
df3 = df2[df2.index.minute.isin([0,30])]
df31 = pd.DataFrame(df3)
df32 = df31.reset_index()
df4 = df1.rolling(1, step=1).agg(['max'])
df5 = df4[df4.index.minute.isin([13,43])]
df51 = pd.DataFrame(df5)
df52 = df51.reset_index()
df6 = pd.concat([df32, df52], axis=1, join='inner')
df7 = pd.DataFrame(df6)
df7['diff_1'] = df7.apply(lambda x: x['temp max'] -
x['temp min'], axis=0)
df7
输出如下，我想从Max中减去最小并显示
ts temp ts temp
min max
0 2023-02-25 00:00:00 0.0 2023-02-25 00:13:00 13.0
1 2023-02-25 00:30:00 68.0 2023-02-25 00:43:00 43.0
2 2023-02-25 01:00:00 60.0 2023-02-25 01:13:00 73.0
分析解答
代码中的错误可能源于agg(['min'])和agg(['max'])操作创建的multi-level列索引。这些列成为层次结构，使它们更加麻烦，可以访问进一步的操作，例如减法。
我没有使用pd.concat，而是利用pd.merge结合了两个数据范围。我还确保temp值是浮动的，以促进数学操作。最后，我直接在合并的dataframe架中计算了差异。
import pandas as pd
# Sample data
ts = ['25/02/2023 0:00', '25/02/2023 0:01', '25/02/2023 0:12', '25/02/2023 0:13']
temp = ['0', '21', '12', '13']
# Create the DataFrame
df = pd.DataFrame(list(zip(ts, temp)), columns=['ts', 'temp'])
df['ts'] = pd.to_datetime(df['ts'])
df['temp'] = df['temp'].astype(float) # Convert to float for calculations
df.set_index('ts', inplace=True)
# Rolling calculations for min and max
df_min = df.rolling(window='30T').min()
df_max = df.rolling(window='30T').max()
# Reset index for merge
df_min.reset_index(inplace=True)
df_max.reset_index(inplace=True)
# Merge the two DataFrames
df_merged = pd.merge(df_min, df_max, on='ts', suffixes=('_min', '_max'))
# Calculate the difference
df_merged['diff'] = df_merged['temp_max'] - df_merged['temp_min']
df_merged
这个适合您吗？
# Sample data
ts = ['25/02/2023 0:00', '25/02/2023 0:01', '25/02/2023 0:12', '25/02/2023 0:13']
temp = ['0', '21', '12', '13']
# Create DataFrame and set index
df = pd.DataFrame(list(zip(ts, temp)), columns=['ts', 'temp'])
df['ts'] = pd.to_datetime(df['ts'])
df['temp'] = df['temp'].astype(float) # Convert to float for calculations
df1 = df.set_index('ts')
# Rolling min and filter for specific minutes
df2 = df1.rolling('1T').min()
df3 = df2[df2.index.minute.isin([0, 30])] # Filtering for minutes 0 and 30
df32 = df3.reset_index()
# Rolling max and filter for specific minutes
df4 = df1.rolling('1T').max()
df5 = df4[df4.index.minute.isin([13, 43])] # Filtering for minutes 13 and 43
df52 = df5.reset_index()
df6 = pd.concat([df32, df52], axis=1, ignore_index=True)
df6.columns = ['ts', 'temp_min', 'ts', 'temp_max'] # ts_min and ts_max are both named 'ts'
# Calculate the difference
df6['diff'] = df6['temp_max'] - df6['temp_min'] # Column name changed to 'diff'
df6