Mayavi中quiver3d()的Colormap

时间:2017-06-28 15:19:23

标签: python python-2.7 plot mayavi

我正在尝试绘制一个矢量场,我正在使用Mayavi2s quiver3d()函数。我希望箭头根据它们在单位球面上的角度着色:

enter image description here

我可以这样做:

def color(theta, phi):
    while(phi < 0):
        phi += 2.0*np.pi
    h = phi / (2.0 * np.pi)
    s = 1
    v = 1.0 - 0.9999999*(theta / np.pi)
    print("h = {}, s = {}, v ={}".format(h,s,v))
    return hsv_to_rgb(h,s,v)

def plot_nice():
    x = np.array([1.,2.,3.])
    y = np.array([0.,0.,0.])
    z = y

    phi   = np.linspace(0,np.pi/2.0,3)
    theta = phi

    u = np.sin(theta) * np.cos(phi)
    v = np.sin(theta) * np.sin(phi)
    w = np.cos(theta)

    obj = quiver3d(x, y, z, u, v, w,
            line_width=3, colormap='hsv', 
            scale_factor=0.8, mode='arrow',resolution=25)

    for i in range(x.shape[0]):
        r,g,b = color(theta[i], phi[i])
        print("R: {}, G: {}, B: {}".format(r,g,b))
        obj = quiver3d(x[i], y[i], z[i], u[i], v[i], w[i],
                line_width=3, color=(r,g,b), colormap='hsv', 
                scale_factor=0.8, mode='arrow',resolution=25)
    return obj

figure(bgcolor=(1,1,1))
plot_nice()

但如果我有数百支箭,那么这很慢。我可以用更快的方式做到这一点:

def plot_fast():
    x = np.array([1.,2.,3.])
    y = np.array([0.,0.,0.])
    z = y

    phi   = np.linspace(0,np.pi/2.0,3)
    theta = phi

    u = np.sin(theta) * np.cos(phi)
    v = np.sin(theta) * np.sin(phi)
    w = np.cos(theta)

    obj = quiver3d(x, y, z, u, v, w,
            line_width=3, colormap='hsv', 
            scale_factor=0.8, mode='arrow',resolution=25)
plot_fast()

但我松开了色彩图: enter image description here

如何创建类似上图的内容,而无需为每个箭头创建新的图?

1 个答案:

答案 0 :(得分:0)

可以通过在事实之后操纵绘图对象来实现。

quiver3d调用,请添加参数scalars=np.mod(phi, 2*np.pi)(例如)和scale_mode='none'。 通话结束后,添加

obj.glyph.color_mode = 'color_by_scalar'

从记忆中,可以单独控制箭头的颜色和大小,但这更精细。