调整散点标记大小以适合imshow单元格大小

时间:2014-08-21 00:47:46

标签: python matplotlib size markers scatter

假设我有一个用imshow绘制的数据,生成如下内容:

X = 10*np.random.rand(5,3)

fig, ax = plt.subplots()
ax.imshow(X, cmap=cm.jet, interpolation='nearest')

enter image description here

我想在此图像上绘制一个散点图,我知道该怎么做。

但是,我希望标记大小完全适合用imshow绘制的数据单元格。即,在前面的例子中,数据是5 x 3.如果我在[0,0]位置上绘制 X 标记,我希望它能够实现橙色单元格区域。

ax.scatter(0,0, marker='x')

enter image description here

我知道scatter有尺码参数,但我不知道如何计算这个尺码。

提前谢谢。

编辑 - 在下面的答案中尝试@Hyry发布的代码。

这是我编写一些要点的代码:

def plot_singularities(x_dim, y_dim, x_steps, y_steps, min_points,
                                              max_points, scalar_field, file_path):
    """
    Plot the singularities of the rotational free scalar field
    :param x_dim : the x dimension of the scalar field
    :param y_dim : the y dimension of the scalar field
    :param x_steps : the discretization in x of the vector field
    :param y_steps : the discretization in y of the vector field
    :param scalar_field : the scalar_field to be plot
    :param min_points : a set (x, y) of min points of the scalar field
    :param max_points : a set (x, y) of max points of the scalar field
    """
    plt.figure()

    ## Plot the scalar field level curves
    x, y = numpy.mgrid[-x_dim/2:x_dim/2:x_steps*1j, -y_dim/2:y_dim/2:y_steps*1j]

    plt.contour(x, y, scalar_field, colors='white', zorder=1, extent=[-x_dim/2.0, x_dim/2.0, -y_dim/2.0, y_dim/2.0])

    ## Plot the critical points
    x = numpy.linspace(-x_dim/2, x_dim/2, x_steps)
    y = numpy.linspace(-y_dim/2, y_dim/2, y_steps)

    # Draw the min points
    x_indices = numpy.nonzero(min_points)[0]
    y_indices = numpy.nonzero(min_points)[1]
    rgba_colors = numpy.zeros((len(x_indices), 4))
    rgba_colors[:, 1] = 0.5
    rgba_colors[:, 3] = min_points[numpy.nonzero(min_points)]
    scatter = plt.scatter(x[x_indices], y[y_indices], color=rgba_colors, marker='$\\otimes$', s=1, zorder=2)
    scatter.__class__ = DataPathCollection

我得到一个更大的标记(绿色标记):

enter image description here

我的数据是100 x 100.标记是否应该变小?

编辑2:

通过一些测试,我发现问题与extent函数中的imshow参数有关:

X = 10 * np.random.rand(5,5)

fig, ax = plt.subplots()
plt.imshow(X, cmap="jet", interpolation='nearest', extent=[-5, 5, -5, 5])
scatter = plt.scatter(0,0, marker='x', s=1)
scatter.__class__ = DataPathCollection
plt.show()

enter image description here

我如何调整代码以克服这种情况?

谢谢。

1 个答案:

答案 0 :(得分:2)

PathCollection的大小在屏幕坐标系中,无法在数据坐标系中设置它。这是一个在数据坐标系中缩放路径的例子。您可以将__class__对象的PathCollection更改为DataPathCollection

交叉标记的原始路径为-0.5~0.5。

from matplotlib.collections import PathCollection
import pylab as pl
import numpy as np

class DataPathCollection(PathCollection):
    def draw(self, renderer):
        if self._sizes is not None:
            affine = self.axes.transData.get_affine()
            m = affine.get_matrix()
            m[:, 2] = 0
            m[2, :] = 0
            self._transforms = [affine.scale(x, x) for x in self._sizes]
        return Collection.draw(self, renderer)    

X = 10*np.random.rand(5,3)

fig, ax = pl.subplots()
ax.imshow(X, cmap="jet", interpolation='nearest')
scatter = ax.scatter(0,0, marker='x', s=1)
scatter.__class__ = DataPathCollection

这是输出:

enter image description here