Python中如何将日期时间设为X轴绘制Y值？日期格式报错求助

Let's break down your error and fix it step by step. That ValueError pops up because of two key issues in your date-parsing logic:

1. Your input date string uses dots as separators (2018.05.01) but your format string %Y%m%d%H:%M expects no separators for the date part.
2. Your date and time are stored in separate columns, but you're trying to parse the date column alone with a format that expects both date and time.

Here are two robust solutions to get your time series plot working:

Solution 1: Use Pandas (Recommended for Simplicity) #

Pandas handles date-time parsing far more smoothly than raw NumPy for this kind of structured data. Here's how to adjust your code:

import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.dates as mdates

# Define column names to match your data structure
column_names = ["date", "time", "open", "high", "low", "close", "volume"]
# Load the CSV file with explicit column names
df = pd.read_csv("DAT_MT_EURUSD_M1_201805.csv", names=column_names)

# Merge date and time columns into a single datetime object
df["datetime"] = pd.to_datetime(
    df["date"] + " " + df["time"],
    format="%Y.%m.%d %H:%M"  # Matches your date-time format exactly
)

# Convert datetime objects to Matplotlib-compatible numeric values
x_values = mdates.date2num(df["datetime"])

# Plot your data (using closing price as an example)
plt.plot(x_values, df["close"])

# Format the X-axis for readability
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter("%Y.%m.%d %H:%M"))
plt.gcf().autofmt_xdate()  # Auto-rotate labels to avoid overlap

plt.xlabel("Datetime")
plt.ylabel("Price")
plt.title("EUR/USD 1-Minute Price Chart")
plt.show()

Solution 2: Custom Converter with NumPy #

If you prefer sticking with NumPy, you'll need a custom converter function that combines the date and time columns before parsing:

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
from datetime import datetime

def parse_datetime(date_bytes, time_bytes):
    # Convert byte strings to regular strings and combine
    datetime_str = f"{date_bytes.decode('utf-8')} {time_bytes.decode('utf-8')}"
    # Parse into a datetime object
    dt = datetime.strptime(datetime_str, "%Y.%m.%d %H:%M")
    # Convert to Matplotlib's numeric date format
    return mdates.date2num(dt)

# Load raw data as byte strings (since we need to process date/time together)
raw_data = np.loadtxt(
    "DAT_MT_EURUSD_M1_201805.csv",
    delimiter=",",
    dtype="S20",
    unpack=False
)

# Extract relevant columns
dates = raw_data[:, 0]
times = raw_data[:, 1]
close_prices = raw_data[:, 5].astype(float)

# Convert date-time pairs to Matplotlib-compatible values
x_values = np.array([parse_datetime(d, t) for d, t in zip(dates, times)])

# Plot the data
plt.plot(x_values, close_prices)
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter("%Y.%m.%d %H:%M"))
plt.gcf().autofmt_xdate()

plt.xlabel("Datetime")
plt.ylabel("Price")
plt.title("EUR/USD 1-Minute Price Chart")
plt.show()

Key Takeaways #

• Always match your format string to the exact structure of your date-time strings: %Y.%m.%d for YYYY.MM.DD, %H:%M for HH:MM.
• When date and time are split across columns, you must combine them first before parsing into a datetime object.
• Pandas is generally the better choice for time series data due to its built-in date handling tools.

内容的提问来源于stack exchange，提问作者Leonid Chebenko