在同一轴上创建两个标签并填充周末间隙

Question

从这个问题中获得了一些指导，在底部创建了我的条形图； Matplotlib - Finance volume overlay

但我似乎无法解决的问题是条形图的空白。正如您在代码中看到的那样，列表y3 有length=21，x、y1、y2 也是如此。空闲的日子是周末，但在我创建任何情节之前，这些日子已从列表中删除 x。即使在我将它们从x-axis 列表中删除之后，我也看不出它们是如何以及为什么会留下空白的。

编辑：我缩短了问题。

from datetime import datetime,timedelta
import matplotlib
import matplotlib.pyplot as plt
import matplotlib.dates as mdates

d = datetime.strptime('2019-02-01','%Y-%m-%d')
#create x-axis which consists of all days in february except weekdays, plus the first day in march
x = []
for i in range(0,29):
    if((d + timedelta(days=i)).date().weekday() > 4):
        continue
    x.append((d + timedelta(days=i)).date())

y1 = [317.92, 319.1, 322.73, 324.22, 322.05, 321.06, 323.79, 325.29, 325.55, 327.9, 329.99, 330.92, 331.33, 332.75, 331.41, 333.22, 332.58, 333.64, 331.19, 329.83, 332.46]
y2 = [-0.377507469, -0.3411700881, 0.055641283900000005, 0.44570105590000003, 0.10930618490000005, -0.2948038151, 0.09190098490000004, 0.12858223490000004, 0.09559385990000004, 0.4806987649, 0.11333709890000004, 0.5247366639000001, 0.11605396190000006, -0.30122159309999996, -0.7405398680999999, -0.7053236033999999, -0.33368165239999986, -0.8210733183999999, -1.2753887724, -1.3106836659999999, -1.3450585547999998]
y3 = [27, 15, 12, 4, 5, 11, 4, 14, 18, 19, 13, 17, 28, 22, 4, 15, 26, 21, 21, 21, 9]

fig, ax1 = plt.subplots()
ax1.plot(x, y1, 'b')
ax1.tick_params('y', colors='b')
#create space at the bottom for the bar-plot
pad = 0.25
yl = ax1.get_ylim()
ax1.set_ylim(yl[0]-(yl[1]-yl[0])*pad,yl[1])

fig.suptitle('Some title', fontsize=16)
ax2 = ax1.twinx()

ax2.plot(x, y2,'r')
ax2.tick_params('y', colors='r')
#create space at the bottom for the bar-plot
pad = 0.25
yl = ax2.get_ylim()
ax2.set_ylim(yl[0]-(yl[1]-yl[0])*pad,yl[1])
#create barplot and make it small by adding a (in this case, this is not a modular solution) max-limit
ax3 = ax1.twinx()
ax3.set_ylim(0,250)
ax3.bar(x, y3, color='green', width=1, alpha=0.5)
ax3.set_xlim(min(x),max(x))
ax3.get_yaxis().set_ticks([])

ax1.set_ylabel('1st Value', color='b')
ax2.set_ylabel('2nd Value', color='r')
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%d-%B'))

plt.show()

Answer 1

您的 x 轴没有周末数据

在以下代码中，您创建了xs，但不包括周末。

x = []
for i in range(0,29):
    if((d + timedelta(days=i)).date().weekday() > 4):
        continue
    x.append((d + timedelta(days=i)).date())

例如，如果 2 月 1 日是星期五，那么 x 数组的第一个元素是 [0, 3, 4, 5]。您没有包括周末的数据，但是通过省略 [1, 2]，您将离开他们占用的空间。

解决方案：去除周末的斑点并处理标签格式

您应该做的是创建没有间隙的代理 x 轴数据，然后重新处理标签。一个完整的解决方案可能如下。

from datetime import datetime,timedelta
import matplotlib
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import matplotlib.ticker as ticker

d = datetime.strptime('2019-02-01','%Y-%m-%d')
#create x-axis which consists of all days in february except weekdays, plus the first day in march
x = []
dates = {}
day = 0
for i in range(0,29):
    if((d + timedelta(days=i)).date().weekday() > 4):
        continue
    x.append(day)
    dates[day] = (d + timedelta(days=i)).date()
    day += 1


def format_date(x, pos=None):
    return dates[int(x)].strftime('%d-%B')

y1 = [317.92, 319.1, 322.73, 324.22, 322.05, 321.06, 323.79, 325.29, 325.55, 327.9, 329.99, 330.92, 331.33, 332.75, 331.41, 333.22, 332.58, 333.64, 331.19, 329.83, 332.46]
y2 = [-0.377507469, -0.3411700881, 0.055641283900000005, 0.44570105590000003, 0.10930618490000005, -0.2948038151, 0.09190098490000004, 0.12858223490000004, 0.09559385990000004, 0.4806987649, 0.11333709890000004, 0.5247366639000001, 0.11605396190000006, -0.30122159309999996, -0.7405398680999999, -0.7053236033999999, -0.33368165239999986, -0.8210733183999999, -1.2753887724, -1.3106836659999999, -1.3450585547999998]
y3 = [27, 15, 12, 4, 5, 11, 4, 14, 18, 19, 13, 17, 28, 22, 4, 15, 26, 21, 21, 21, 9]

fig, ax1 = plt.subplots()
ax1.plot(x, y1, 'b')
ax1.tick_params('y', colors='b')
#create space at the bottom for the bar-plot
pad = 0.25
yl = ax1.get_ylim()
ax1.set_ylim(yl[0]-(yl[1]-yl[0])*pad,yl[1])

fig.suptitle('Some title', fontsize=16)
ax2 = ax1.twinx()

ax2.plot(x, y2,'r')
ax2.tick_params('y', colors='r')
#create space at the bottom for the bar-plot
pad = 0.25
yl = ax2.get_ylim()
ax2.set_ylim(yl[0]-(yl[1]-yl[0])*pad,yl[1])
#create barplot and make it small by adding a (in this case, this is not a modular solution) max-limit
ax3 = ax1.twinx()
ax3.set_ylim(0,250)
ax3.bar(x, y3, color='green', width=1, alpha=0.5)
ax3.set_xlim(min(x),max(x))
ax3.get_yaxis().set_ticks([])

ax1.set_ylabel('1st Value', color='b')
ax2.set_ylabel('2nd Value', color='r')
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(ticker.FuncFormatter(format_date))

plt.show()

这会产生图像

这是相同的数据，但没有周末间隔。

差异说明

我为这个练习改变了三个方面。首先，我设置x 最终代表数组[0, 1, 2, ..., 20]，并创建一个字典dates 来跟踪x[i] 代表哪一天。

x = []
dates = {}
day = 0
for i in range(0,29):
    if((d + timedelta(days=i)).date().weekday() > 4):
        continue
    x.append(day)
    dates[day] = (d + timedelta(days=i)).date()
    day += 1

然后我创建了一个函数 format_date 来提供给 matplotlib.ticker 来处理 x 轴的标签。

def format_date(x, pos=None):
    return dates[int(x)].strftime('%d-%B')

最后，我将x轴设置为使用此格式功能。这里使用matplotlib.ticker，必须导入。

# in the imports, add
import matplotlib.ticker as ticker

# when describing the x-axis, use
plt.gca().xaxis.set_major_formatter(ticker.FuncFormatter(format_date))