python - 在 matplotlib 3d 散点图中更改数据点的颜色并通过按键将其删除

Question

我在 matplotlib 中有一个 3d 散点图，并设置了注释，灵感来自答案 here ，特别是 Don Cristobal .

我已经设置了一些基本的事件捕获代码，但经过几天的尝试，我仍然没有实现我的目标。它们是:

(i) 使用鼠标左键选择时，将点(点)的颜色从蓝色更改为例如深蓝色/绿色。

(ii) 在按下'delete'键后删除在(i)中选择的任何选定点，包括任何注释

(iii) 使用 selection rectangle 在 (i) 中选择多个点并使用“删除”键删除

我尝试了很多方法，包括动画图表以根据数据变化进行更新、操纵艺术家参数、通过例如更改数据点xs, ys, zs = graph._offsets3d(似乎没有记录)，但无济于事。

我尝试在 onpick(event) 函数中:

(i) 通过 event.ind 与点交互以使用 event.artist.set_face_colour() 更改颜色

(ii) 使用 artist.remove() 删除点

(iii) 使用 xs, ys, zs = graph._offsets3d 删除点，通过索引 (event.ind[0]) 从 xs、ys 和 zs 中删除相关点，然后通过 graph._offsets3d 重置图形点= xs_new, ys_new, zs_new

(iv) 重绘图表，或仅重绘图表的相关部分(blitting？)

没有成功!

我目前的代码大致如下。事实上，我有几百个点，而不是下面简化示例中的 3 个。如果可能的话，我希望图表能够顺利更新，尽管只是得到可用的东西会很棒。执行此操作的大部分代码可能应该位于“onpick”中，因为这是处理拾取事件的函数(参见 event handler )。我保留了一些我的代码尝试，注释掉了，我希望这可能会有一些用处。 “forceUpdate”函数旨在在事件触发时更新图形对象，但我不相信它目前有任何作用。函数 on_key(event) 目前似乎也不起作用:大概必须有一个设置才能确定要删除的点，例如所有具有已更改的默认面色的艺术家(例如，删除所有颜色为深蓝色/绿色而不是浅蓝色的点)。

非常感谢任何帮助。

代码(如下)被调用:

visualize3DData (Y, ids, subindustry)

一些示例数据点如下:

#Datapoints
Y = np.array([[ 4.82250000e+01,  1.20276889e-03,  9.14501289e-01], [ 6.17564688e+01,  5.95020883e-02, -1.56770827e+00], [ 4.55139000e+01,  9.13454423e-02, -8.12277299e+00]])

#Annotations
ids = ['a', 'b', 'c']

subindustry =  'example'

我当前的代码在这里:

import matplotlib.pyplot as plt, numpy as np
from mpl_toolkits.mplot3d import proj3d

def visualize3DData (X, ids, subindus):
    """Visualize data in 3d plot with popover next to mouse position.

    Args:
        X (np.array) - array of points, of shape (numPoints, 3)
    Returns:
        None
    """
    fig = plt.figure(figsize = (16,10))
    ax = fig.add_subplot(111, projection = '3d')
    graph  = ax.scatter(X[:, 0], X[:, 1], X[:, 2], depthshade = False, picker = True)  

    def distance(point, event):
        """Return distance between mouse position and given data point

        Args:
            point (np.array): np.array of shape (3,), with x,y,z in data coords
            event (MouseEvent): mouse event (which contains mouse position in .x and .xdata)
        Returns:
            distance (np.float64): distance (in screen coords) between mouse pos and data point
        """
        assert point.shape == (3,), "distance: point.shape is wrong: %s, must be (3,)" % point.shape

        # Project 3d data space to 2d data space
        x2, y2, _ = proj3d.proj_transform(point[0], point[1], point[2], plt.gca().get_proj())
        # Convert 2d data space to 2d screen space
        x3, y3 = ax.transData.transform((x2, y2))

        return np.sqrt ((x3 - event.x)**2 + (y3 - event.y)**2)


    def calcClosestDatapoint(X, event):
        """"Calculate which data point is closest to the mouse position.

        Args:
            X (np.array) - array of points, of shape (numPoints, 3)
            event (MouseEvent) - mouse event (containing mouse position)
        Returns:
            smallestIndex (int) - the index (into the array of points X) of the element closest to the mouse position
        """
        distances = [distance (X[i, 0:3], event) for i in range(X.shape[0])]
        return np.argmin(distances)


    def annotatePlot(X, index, ids):
        """Create popover label in 3d chart

        Args:
            X (np.array) - array of points, of shape (numPoints, 3)
            index (int) - index (into points array X) of item which should be printed
        Returns:
            None
        """
        # If we have previously displayed another label, remove it first
        if hasattr(annotatePlot, 'label'):
            annotatePlot.label.remove()
        # Get data point from array of points X, at position index
        x2, y2, _ = proj3d.proj_transform(X[index, 0], X[index, 1], X[index, 2], ax.get_proj())
        annotatePlot.label = plt.annotate( ids[index],
            xy = (x2, y2), xytext = (-20, 20), textcoords = 'offset points', ha = 'right', va = 'bottom',
            bbox = dict(boxstyle = 'round,pad=0.5', fc = 'yellow', alpha = 0.5),
            arrowprops = dict(arrowstyle = '->', connectionstyle = 'arc3,rad=0'))
        fig.canvas.draw()


    def onMouseMotion(event):
        """Event that is triggered when mouse is moved. Shows text annotation over data point closest to mouse."""
        closestIndex = calcClosestDatapoint(X, event)
        annotatePlot (X, closestIndex, ids) 


    def onclick(event):
        print('%s click: button=%d, x=%d, y=%d, xdata=%f, ydata=%f' %
              ('double' if event.dblclick else 'single', event.button,
               event.x, event.y, event.xdata, event.ydata))

    def on_key(event):
        """
        Function to be bound to the key press event
        If the key pressed is delete and there is a picked object,
        remove that object from the canvas
        """
        if event.key == u'delete':
            ax = plt.gca()
            if ax.picked_object:
                ax.picked_object.remove()
                ax.picked_object = None
                ax.figure.canvas.draw()

    def onpick(event):

        xmouse, ymouse = event.mouseevent.xdata, event.mouseevent.ydata
        artist = event.artist
        # print(dir(event.mouseevent))
        ind = event.ind
        # print('Artist picked:', event.artist)
        # # print('{} vertices picked'.format(len(ind)))
        print('ind', ind)
        # # print('Pick between vertices {} and {}'.format(min(ind), max(ind) + 1))
        # print('x, y of mouse: {:.2f},{:.2f}'.format(xmouse, ymouse))
        # # print('Data point:', x[ind[0]], y[ind[0]])
        #
        # # remove = [artist for artist in pickable_artists if     artist.contains(event)[0]]
        # remove = [artist for artist in X if artist.contains(event)[0]]
        #
        # if not remove:
        #     # add a pt
        #     x, y = ax.transData.inverted().transform_point([event.x,     event.y])
        #     pt, = ax.plot(x, y, 'o', picker=5)
        #     pickable_artists.append(pt)
        # else:
        #     for artist in remove:
        #         artist.remove()
        # plt.draw()
        # plt.draw_idle()

        xs, ys, zs = graph._offsets3d
        print(xs[ind[0]])
        print(ys[ind[0]])
        print(zs[ind[0]])
        print(dir(artist))

        # xs[ind[0]] = 0.5
        # ys[ind[0]] = 0.5
        # zs[ind[0]] = 0.5   
        # graph._offsets3d = (xs, ys, zs)

        # print(artist.get_facecolor())
        # artist.set_facecolor('red')
        graph._facecolors[ind, :] = (1, 0, 0, 1)

        plt.draw()

    def forceUpdate(event):
        global graph
        graph.changed()

    fig.canvas.mpl_connect('motion_notify_event', onMouseMotion)  # on mouse motion    
    fig.canvas.mpl_connect('button_press_event', onclick)
    fig.canvas.mpl_connect('pick_event', onpick)
    fig.canvas.mpl_connect('draw_event', forceUpdate)

    plt.tight_layout()

    plt.show()

Answer 1

好的，我至少为您提供了部分解决方案，没有矩形选择，但您可以选择多个点并使用一个 key_event 删除它们。

要改变颜色，你需要改变graph._facecolor3d，提示在this中。关于 set_facecolor 未设置 _facecolor3d 的错误报告。

将您的函数重写为一个类以摆脱任何需要的 global 变量也可能是个好主意。

我的解决方案有一些不太漂亮的部分，我需要在删除数据点后重新绘制图形，我无法删除和更新工作。还有 (见下面的编辑 2)。我还没有实现如果删除最后一个数据点会发生什么。

您的函数 on_key(event) 不起作用的原因很简单:您忘记连接它。

所以这是一个应该满足目标 (i) 和 (ii) 的解决方案:

import matplotlib.pyplot as plt, numpy as np
from mpl_toolkits.mplot3d import proj3d

class Class3DDataVisualizer:    
    def __init__(self, X, ids, subindus, drawNew = True):

        self.X = X;
        self.ids = ids
        self.subindus = subindus

        self.disconnect = False
        self.ind = []
        self.label = None

        if drawNew:        
            self.fig = plt.figure(figsize = (7,5))
        else:
            self.fig.delaxes(self.ax)
        self.ax = self.fig.add_subplot(111, projection = '3d')
        self.graph  = self.ax.scatter(self.X[:, 0], self.X[:, 1], self.X[:, 2], depthshade = False, picker = True, facecolors=np.repeat([[0,0,1,1]],X.shape[0], axis=0) )         
        if drawNew and not self.disconnect:
            self.fig.canvas.mpl_connect('motion_notify_event', lambda event: self.onMouseMotion(event))  # on mouse motion    
            self.fig.canvas.mpl_connect('pick_event', lambda event: self.onpick(event))
            self.fig.canvas.mpl_connect('key_press_event', lambda event: self.on_key(event))

        self.fig.tight_layout()
        self.fig.show()


    def distance(self, point, event):
        """Return distance between mouse position and given data point

        Args:
            point (np.array): np.array of shape (3,), with x,y,z in data coords
            event (MouseEvent): mouse event (which contains mouse position in .x and .xdata)
        Returns:
            distance (np.float64): distance (in screen coords) between mouse pos and data point
        """
        assert point.shape == (3,), "distance: point.shape is wrong: %s, must be (3,)" % point.shape

        # Project 3d data space to 2d data space
        x2, y2, _ = proj3d.proj_transform(point[0], point[1], point[2], plt.gca().get_proj())
        # Convert 2d data space to 2d screen space
        x3, y3 = self.ax.transData.transform((x2, y2))

        return np.sqrt ((x3 - event.x)**2 + (y3 - event.y)**2)


    def calcClosestDatapoint(self, event):
        """"Calculate which data point is closest to the mouse position.

        Args:
            X (np.array) - array of points, of shape (numPoints, 3)
            event (MouseEvent) - mouse event (containing mouse position)
        Returns:
            smallestIndex (int) - the index (into the array of points X) of the element closest to the mouse position
        """
        distances = [self.distance (self.X[i, 0:3], event) for i in range(self.X.shape[0])]
        return np.argmin(distances)


    def annotatePlot(self, index):
        """Create popover label in 3d chart

        Args:
            X (np.array) - array of points, of shape (numPoints, 3)
            index (int) - index (into points array X) of item which should be printed
        Returns:
            None
        """
        # If we have previously displayed another label, remove it first
        if self.label is not None:
            self.label.remove()
        # Get data point from array of points X, at position index
        x2, y2, _ = proj3d.proj_transform(self.X[index, 0], self.X[index, 1], self.X[index, 2], self.ax.get_proj())
        self.label = plt.annotate( self.ids[index],
            xy = (x2, y2), xytext = (-20, 20), textcoords = 'offset points', ha = 'right', va = 'bottom',
            bbox = dict(boxstyle = 'round,pad=0.5', fc = 'yellow', alpha = 0.5),
            arrowprops = dict(arrowstyle = '->', connectionstyle = 'arc3,rad=0'))
        self.fig.canvas.draw()


    def onMouseMotion(self, event):
        """Event that is triggered when mouse is moved. Shows text annotation over data point closest to mouse."""
        closestIndex = self.calcClosestDatapoint(event)
        self.annotatePlot (closestIndex) 


    def on_key(self, event):
        """
        Function to be bound to the key press event
        If the key pressed is delete and there is a picked object,
        remove that object from the canvas
        """
        if event.key == u'delete':
            if self.ind:
                self.X = np.delete(self.X, self.ind, axis=0)
                self.ids = np.delete(ids, self.ind, axis=0)
                self.__init__(self.X, self.ids, self.subindus, False)
            else:
                print('nothing selected')

    def onpick(self, event):
        self.ind.append(event.ind)
        self.graph._facecolor3d[event.ind] = [1,0,0,1]



#Datapoints
Y = np.array([[ 4.82250000e+01,  1.20276889e-03,  9.14501289e-01], [ 6.17564688e+01,  5.95020883e-02, -1.56770827e+00], [ 4.55139000e+01,  9.13454423e-02, -8.12277299e+00], [3,  8, -8.12277299e+00]])
#Annotations
ids = ['a', 'b', 'c', 'd']

subindustries =  'example'

Class3DDataVisualizer(Y, ids, subindustries)

要实现您的矩形选择，您必须覆盖当前在拖动(旋转 3D 图)期间发生的情况，或者更简单的解决方案是通过连续两次点击来定义您的矩形。

然后使用 proj3d.proj_transform 查找矩形内的数据，找到所述数据的索引并使用 self.graph._facecolor3d[idx] 重新着色> 函数并用这些索引填充 self.ind，之后点击删除将负责删除 self.ind 指定的所有数据。

编辑: 我在 __init__ 中添加了两行，删除了 ax/subplot，然后在删除数据点后添加新的。我注意到在删除了几个数据点后绘图交互变得缓慢，因为该图只是在每个子图中绘制。

编辑 2: 我发现了如何修改数据而不是重新绘制整个图，如 this answer 中所述你将不得不修改 _offsets3d，它奇怪地为 x 和 y 返回一个元组，但是为 z 返回一个数组.

你可以修改它使用

(x,y,z) = self.graph._offsets3d # or event.artist._offsets3d
xNew = x[:int(idx)] + x[int(idx)+1:]
yNew = y[:int(idx)] + y[int(idx)+1:]
z = np.delete(z, int(idx))
self.graph._offsets3d = (xNew,yNew,z) # or event.artist._offsets3d

但是你会在循环中删除多个数据点时遇到问题，因为你之前存储的索引在第一个循环后将不再适用，你必须更新 _facecolor3d，标签列表...所以我选择保留答案原样，因为用新数据重新绘制绘图似乎比这更容易和更清晰。