求助：日期格式无法处理及'int'对象无'toordinal'属性报错

Hey there, let's break down and solve your two problems one by one:

1. Why you're seeing "Date format can't be handled as date" #

Your df1['Date'] series contains float values, which pandas/matplotlib can't natively interpret as dates. Floats (like 20230101.0) aren't recognized as valid date inputs, hence the format error.

2. Why the AttributeError: 'int' object has no attribute 'toordinal' happens #

When you try to plot with float dates, matplotlib implicitly converts them to integers (to drop the decimal) and then tries to treat those integers as date objects. But plain integers don't have the toordinal() method that matplotlib uses to handle date plotting—this is exactly what triggers the error.

Step-by-Step Solutions #

First: Convert the Date column to a proper datetime type

You'll need to convert those float values into pandas datetime objects. The exact code depends on your date's underlying format, but here's the most common scenario (assuming your floats represent dates like YYYYMMDD, e.g., 20230101.0):

import pandas as pd

# Convert float -> int (to remove .0) -> string -> datetime
df1['Date'] = pd.to_datetime(df1['Date'].astype(int).astype(str), format='%Y%m%d')

• If your date format is different: Adjust the format parameter. For example, if your floats are 2023.01.01, you'd first replace the dots with hyphens:

  df1['Date'] = pd.to_datetime(
      df1['Date'].astype(str).str.replace('.', '-'),
      format='%Y-%m-%d'
  )
  

• Handle missing values: If there are NaNs in your Date column, add errors='coerce' to convert invalid entries to NaT (Not a Time):

  df1['Date'] = pd.to_datetime(df1['Date'].astype(int).astype(str), format='%Y%m%d', errors='coerce')
  

Second: Update your plotting code

Once your Date column is in datetime format, matplotlib will handle it correctly without the toordinal error. Here's your full corrected code (I also added an initial value for performance since it looks like you're starting from index 2265):

import pandas as pd
import matplotlib.pyplot as plt

# Fix the Date column first
df1['Date'] = pd.to_datetime(df1['Date'].astype(int).astype(str), format='%Y%m%d')

# Initialize performance array with matching length to df1
performance = [0.0] * df1.shape[0]
performance[2265] = 1.0  # Set your starting performance value here

# Calculate performance
for i in range(2266, df1.shape[0]):
    performance[i] = performance[i-1] * (1 + r[i])

df1['performance'] = performance
print(performance[2267])

# Plot without date errors
plt.plot(df1['Date'][2265:4616], df1['performance'][2265:4616])
plt.xlabel('Date')
plt.ylabel('Performance')
plt.xticks(rotation=45)  # Optional: Rotate dates for readability
plt.show()

内容的提问来源于stack exchange，提问作者Rawia Sammout