向散点图添加图例

Question

这个问题已经被问到了，但我想找到一个更清楚的解决办法。

给定X是100x2数据，标签是标签向量（从1到9），我绘制散点图如下：

pl.scatter(X[:,0], X[:,1], c = labels)
pl.show()

如何添加图例来解释一行代码中的颜色？其他解决方案分别绘制每个标签：

a = pl.scatter(X1[:,0], X1[:,1], color = "red")
b = pl.scatter(X2[:,0], X2[:,1], color = "green")
c = pl.scatter(X3[:,0], X3[:,1], color = "blue")
pl.legend((a,b,c), ("line 1", "line 2", "line 3")
pl.show()

Answer 1

只需标记每个绘图并调用图例（），就像您所做的那样：）

plt.scatter(x1,y1,label=str(pointset1))
plt.scatter(x2,y2,label=str(pointset2))
plt.scatter(x3,y3,label=str(pointset3))

plt.legend(loc='upper right', numpoints=1, ncol=3, fontsize=8, bbox_to_anchor=(1,1))
plt.show()

Answer 2

可以使用所需图例创建虚拟散点图，如下所示：

pl.scatter(X[:,0], X[:,1], c = labels)

for item in labels:
    #dummy plot just to create the legend
    pl.scatter([], [], c = item, label = item)
#loc = 0 is for the best position of the legend
#scatterpoints = 1 will only show one point in the legend instead of multiple points
plt.legend(loc = 0, scatterpoints = 1)
pl.show()