使用matplotlib降序x轴值

Question

我想画一个柱状图，x轴的数值要从高到低排列。

举个例子：

为了说明问题，我的图是这样画的：

我需要这个图的顺序是：星期一（最高值）、星期三、星期二（最低值）这样排列。

这是我目前的代码：

x_axis = ['a','b','c'...'z']
y_axis = [#...#...#] number values for each letter in xaxis

def barplot(x_axis, y_axis): #x and y axis defined in another function
    x_label_pos = range(len(y_axis))
    plot.bar(x_label_pos, y_axis)
    plot.yticks(range(0, int(max(y_axis) + 2), 2))
    plot.xticks(x_axis) 

Answer 1

下面是一个简单的例子，可以满足你的需求。其实你的问题和matplotlib没有关系，主要是需要重新整理一下你的输入数据，让它按照你想要的顺序排列。

import matplotlib.pyplot as plt

# some dummy lists with unordered values 
x_axis = ['a','b','c']
y_axis = [1,3,2]

def barplot(x_axis, y_axis): 
    # zip the two lists and co-sort by biggest bin value         
    ax_sort = sorted(zip(y_axis,x_axis), reverse=True)
    y_axis = [i[0] for i in ax_sort]
    x_axis = [i[1] for i in ax_sort]

    # the above is ugly and would be better served using a numpy recarray

    # get the positions of the x coordinates of the bars
    x_label_pos = range(len(x_axis))

    # plot the bars and align on center of x coordinate
    plt.bar(x_label_pos, y_axis,align="center")

    # update the ticks to the desired labels
    plt.xticks(x_label_pos,x_axis)


barplot(x_axis, y_axis)
plt.show()

Answer 2

# grab a reference to the current axes
ax = plt.gca()
# set the xlimits to be the reverse of the current xlimits
ax.set_xlim(ax.get_xlim()[::-1])
# call `draw` to re-render the graph
plt.draw()

matplotlib 会自动处理你设置的坐标轴范围，如果你把x轴的左边界设置得比右边界大（y轴也是一样），它会“聪明地”调整显示的内容。