matplotlib命令与格式之tick坐标轴日期格式(设置日期主副刻度)
1.横坐标设置时间格式
import matplotlib.pyplot as plt import matplotlib.dates as mdates # 配置横坐标为日期格式 plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%Y/%m/%d')) plt.gca().xaxis.set_major_locator(mdates.DayLocator())
例子:
from datetime import datetime import matplotlib.dates as mdates import matplotlib.pyplot as plt # 生成横纵坐标信息 dates = ['01/02/1991', '01/03/1991', '01/04/1991'] xs = [datetime.strptime(d, '%m/%d/%Y').date() for d in dates] ys = range(len(xs)) # 配置横坐标 plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%m/%d/%Y')) plt.gca().xaxis.set_major_locator(mdates.DayLocator()) # Plot plt.plot(xs, ys) plt.gcf().autofmt_xdate() # 自动旋转日期标记 plt.show()
2.设置日期坐标轴主副刻度值
所有坐标轴日期格式类型
MinuteLocator: locate minutes(f)
HourLocator: locate hours
DayLocator: locate specified days of the month
WeekdayLocator: Locate days of the week, e.g., MO, TU
MonthLocator: locate months, e.g., 7 for july
YearLocator: locate years that are multiples of base
RRuleLocator: locate using a matplotlib.dates.rrulewrapper. The rrulewrapper is a simple wrapper around adateutil.rrule (dateutil) which allow almost arbitrary date tick specifications. See rrule example.
AutoDateLocator: On autoscale, this class picks the best MultipleDateLocator to set the view limits and the tick locations.
(1)获取坐标轴日期格式类型
from matplotlib.dates import DateFormatter, WeekdayLocator, DayLocator, MONDAY,YEARLY #获取每月一日数据 monthdays = MonthLocator() #获取每周一的日期数据 mondays = WeekdayLocator(MONDAY) #获取每日数据 alldays = DayLocator() # import constants for the days of the week from matplotlib.dates import MO, TU, WE, TH, FR, SA, SU # tick on mondays every week loc = WeekdayLocator(byweekday=MO, tz=tz) # tick on mondays and saturdays loc = WeekdayLocator(byweekday=(MO, SA)) # tick on mondays every second week loc = WeekdayLocator(byweekday=MO, interval=2) # tick every 5th easter(每隔5个选1个) rule = rrulewrapper(YEARLY, byeaster=1, interval=5) loc = RRuleLocator(rule)
(2)设置坐标轴日期格式
#设置主副刻度 ax.xaxis.set_major_locator(mondays)ax.xaxis.set_minor_locator(alldays) #设置坐标轴刻度标签格式 mondayFormatter = DateFormatter('%Y-%m-%d') # 如:2-29-2015dayFormatter = DateFormatter('%d') # 如:12ax.xaxis.set_major_formatter(mondayFormatter) #字符串旋转 for label in ax1.get_xticklabels(): label.set_rotation(30) label.set_horizontalalignment('right')
(3)例子
import matplotlib.pyplot as plt import matplotlib.dates as mdates from datetime import datetime #销售数据 dates=[20171101,20171102,20171103,20171104] sales=[102.1,100.6,849,682] #将dates改成日期格式 x= [datetime.strptime(str(d), '%Y%m%d').date() for d in dates] #figure布局 fig=plt.figure(figsize=(8,4)) ax1=fig.add_subplot(1,1,1) #绘图 ax1.plot(x,y,ls='--',lw=3,color='b',marker='o',ms=6, mec='r',mew=2, mfc='w',label='业绩趋势走向') plt.gcf().autofmt_xdate() # 自动旋转日期标记 #设置x轴主刻度格式 alldays = mdates.DayLocator() #主刻度为每天 ax1.xaxis.set_major_locator(alldays) #设置主刻度 ax1.xaxis.set_major_formatter(mdates.DateFormatter('%Y%m%d')) #设置副刻度格式 hoursLoc = mpl.dates.HourLocator(interval=6) #为6小时为1副刻度 ax1.xaxis.set_minor_locator(hoursLoc) ax1.xaxis.set_minor_formatter(mdates.DateFormatter('%H')) #参数pad用于设置刻度线与标签间的距离 ax1.tick_params(pad=10) #显示图像 plt.show()
3.设置日期时间刻度值
import matplotlib.pyplot as plt import numpy as np import matplotlib as mpl import datetime as dt fig = plt.figure() ax2 = fig.add_subplot(212) date2_1 = dt.datetime(2008,9,23) date2_2 = dt.datetime(2008,10,3) delta2 = dt.timedelta(days=1) dates2 = mpl.dates.drange(date2_1, date2_2, delta2) y2 = np.random.rand(len(dates2)) ax2.plot_date(dates2, y2, linestyle='-') dateFmt = mpl.dates.DateFormatter('%Y-%m-%d') ax2.xaxis.set_major_formatter(dateFmt) daysLoc = mpl.dates.DayLocator() hoursLoc = mpl.dates.HourLocator(interval=6) ax2.xaxis.set_major_locator(daysLoc) ax2.xaxis.set_minor_locator(hoursLoc) fig.autofmt_xdate(bottom=0.18) fig.subplots_adjust(left=0.18) ax1 = fig.add_subplot(211) date1_1 = dt.datetime(2008, 9, 23) date1_2 = dt.datetime(2009, 2, 16) delta1 = dt.timedelta(days=10) dates1 = mpl.dates.drange(date1_1, date1_2, delta1) y1 = np.random.rand(len(dates1)) ax1.plot_date(dates1, y1, linestyle='--') monthsLoc = mpl.dates.MonthLocator() weeksLoc = mpl.dates.WeekdayLocator() ax1.xaxis.set_major_locator(monthsLoc) ax1.xaxis.set_minor_locator(weeksLoc) monthsFmt = mpl.dates.DateFormatter('%b') ax1.xaxis.set_major_formatter(monthsFmt) plt.show()
以上就是本文的全部内容,希望对大家的学习有所帮助,也希望大家多多支持【听图阁-专注于Python设计】。