如何在X轴上相对于Y值绘制日期和时间（Python）

Question

Data

2018.05.01,01:15,1.206870,1.206920,1.206870,1.206920,0
2018.05.01,01:16,1.206910,1.206910,1.206810,1.206820,0

Code

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

data1 = pd.read_csv('DAT_MT_EURUSD_M1_201805.csv')

date, time, closep, highp, openp = np.loadtxt(data1, 
                                              delimiter=',',
                                              unpack = True,
                                              dtype=float,
                                              converters ={0: bytespdate2num('%Y%m%d%H:%M')})

At the end have an error

ValueError: time data u'2018.05.01' does not match format '%Y%m%d%H:%M'

It seems like i have to write a converter function to fit the proper date format, but I don't know how.

Please suggest.

Answer 1

Just change the format to %Y.%m.%d should do the trick, you're missing the dots in between.

If not you can also try dateutil which parses most of the date formats for you:

import dateutil

dateutil.parser.parse(my_date_str)

Answer 2

Your date and time are separated by a comma, which you are splitting on, so you are getting incorrect results. You should pre-format your file to remove that first comma:

data = []
with open('test.txt') as f:
  for line in f:
    data.append(line.replace(',', ' ', 1))

with open('test.txt', 'w') as outf:
  outf.write(''.join(data))

# 2018.05.01 01:15,1.206870,1.206920,1.206870,1.206920,0
# 2018.05.01 01:16,1.206910,1.206910,1.206810,1.206820,0

Make sure you only run this once or else more commas will be removed.

Which can then easily be parsed into a time using pandas (which you're already using):

df = pd.read_csv('test.txt', header=None)
pd.to_datetime(df[0])

0   2018-05-01 01:15:00
1   2018-05-01 01:16:00
Name: 0, dtype: datetime64[ns]

Answer 3

You can directly ask pd.read_csv() to parse the date and time with the parse_date keyword argument:

import pandas as pd
import matplotlib.pyplot as plt

# Read data AND parse time stamps (date + time) given by columns 0 and 1
data = pd.read_csv('data.csv', sep=',', header=None, parse_dates=[0, 1])

# Plot and save figure
plt.plot(data[[1]], data[[2, 3, 4, 5]])
plt.savefig('pandas_plot.png')

Given the two lines of data you provided (that I stored in the data.csv file), this script produces the following figure (where you can actually see your data sampled at two different dates):