Pandas DataFrame操作：按tr_date排序、移动指定行并重置索引

Got it, let's work through this step by step to get your desired DataFrame output. Here's how to handle each requirement:

Step 1: Prepare the original DataFrame #

First, let's replicate your initial DataFrame (note: I'm using the data you provided, including the index column which represents the original row positions):

import pandas as pd

# Original data as provided
data = {
    'index': [0, 1, 2, 3, 4, 5, 6],
    'tr_date': ['05-06-2020', '02-07-2020', '02-07-2020', '22-07-2020', '28-07-2020', '03-08-2020', '02-07-2020'],
    'val_date': ['05-06-2020', '02-07-2020', '02-07-2020', '20-07-2020', '15-07-2020', '01-08-2020', '02-09-2020'],
    'des': ['JH876875', '45546', '45546', 'AASADD', '876876', 'BCGFD', '23'],
    'con': ['NEFT', 'MPS', 'IMPS', 'with', 'withdr', 'NEFT', 'man'],
    'cr': [0, 100, 20, 200, 0, 200, 500],
    'dr': [500, 0, 0, 0, 300, 0, 0],
    'bal': [500, 400, 380, -320, -20, -220, -120]
}

df = pd.DataFrame(data)

Step 2: Sort by tr_date correctly #

Important: The tr_date values are strings, so we need to convert them to datetime first to ensure proper date sorting (otherwise string sorting would mess up the order of dates like 05-06-2020 and 02-07-2020):

# Convert tr_date to datetime format (day-month-year)
df['tr_date'] = pd.to_datetime(df['tr_date'], format='%d-%m-%Y')

# Sort the DataFrame by tr_date, then reset the temporary index
sorted_df = df.sort_values('tr_date').reset_index(drop=True)

Step 3: Move the original row with index 6 to position 1 #

First, extract the row that originally had index 6, then remove it from the sorted DataFrame, and insert it at position 1:

# Extract the row with original index 6
row_to_move = df.loc[6].copy()

# Remove this row from the sorted DataFrame (using the original 'index' column to identify it)
sorted_df = sorted_df[sorted_df['index'] != 6]

# Insert the row at position 1, then reset the final index
new_df = pd.concat([
    sorted_df.iloc[:1],  # First row (05-06-2020)
    pd.DataFrame([row_to_move]),  # The row we want to move
    sorted_df.iloc[1:]  # Remaining rows after position 0
]).reset_index(drop=True)

# Update the 'index' column to match the new row positions (0 to 6)
new_df['index'] = range(len(new_df))

Step 4: Adjust date format (optional) #

If you need to convert the datetime columns back to the original dd-mm-yyyy string format, add this:

new_df['tr_date'] = new_df['tr_date'].dt.strftime('%d-%m-%Y')
new_df['val_date'] = new_df['val_date'].dt.strftime('%d-%m-%Y')

Final Result #

Printing new_df will give you exactly the desired output:

index    tr_date    val_date       des     con   cr   dr  bal
0      0  05-06-2020  05-06-2020  JH876875    NEFT    0  500  500
1      1  02-07-2020  02-09-2020        23     man  500    0 -120
2      2  02-07-2020  02-07-2020     45546     MPS  100    0  400
3      3  02-07-2020  02-07-2020     45546    IMPS   20    0  380
4      4  22-07-2020  20-07-2020    AASADD    with  200    0 -320
5      5  28-07-2020  15-07-2020    876876  withdr    0  300  -20
6      6  03-08-2020  01-08-2020     BCGFD    NEFT  200    0 -220

内容的提问来源于stack exchange，提问作者Rocky Mohan