使用 matplotlib 制作 3D 动画，连接点以创建移动的简笔图

Question

I am currently having some trouble with my code which animates some time-series data, and I cannot quite figure it out. Basically I have 12 tags which I am animating through time. Each tag has a trajectory in time such that the movement path can be seen for each tag as it progresses (have a look at the image attached). Now I would like the animation to also include the lines between pairs of tags (ie pairs of points - for example, how to add an animation line between the yellow and green tags), but I am not entirely sure how to do this. This is code adapted from jakevdp.github.io.

Here is the code thus far.

"""
Full animation of a walking event (note: a lot of missing data) 
"""
import numpy as np
import pandas as pd
import matplotlib
matplotlib.use('TkAgg') # Need to use in order to run on mac
from matplotlib import pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from matplotlib.colors import cnames
from matplotlib import animation

#=============================================================================================

t_start = 1917 # start frame
t_end = 2130 # end frame

data = pd.read_csv('~/Smart-first_phase_NaN-zeros.csv') # only coordinate data
df = data.loc[t_start:t_end,'Shoulder_left_x':'Ankle_right_z']

# Find max and min values for animation ranges
df_minmax = pd.DataFrame(index=list('xyz'),columns=range(2))
for i in list('xyz'):
    c_max = df.filter(regex='_{}'.format(i)).max().max()
    c_min = df.filter(regex='_{}'.format(i)).min().min()
    df_minmax.ix[i] = np.array([c_min,c_max])

df_minmax = 1.3*df_minmax # increase by 30% to make animation look better

df.columns  = np.repeat(range(12),3) # store cols like this for simplicity
N_tag = df.shape[1]/3 # nr of tags used (all)

N_trajectories = N_tag

t = np.linspace(0,data.Time[t_end],df.shape[0]) # pseudo time-vector for first walking activity
x_t = np.zeros(shape=(N_tag,df.shape[0],3)) # empty animation array (3D)

for tag in range(12):
    # store data in numpy 3D array: (tag,time-stamp,xyz-coordinates)
    x_t[tag,:,:] = df[tag]


#===STICK-LINES========================================================================================
#xx = [x_t[1,:,0],x_t[2,:,0]]
#yy = [x_t[1,:,1],x_t[2,:,1]]
#zz = [x_t[1,:,2],x_t[2,:,2]] 
#======================================================================================================

# Set up figure & 3D axis for animation
fig = plt.figure()
ax = fig.add_axes([0, 0, 1, 1], projection='3d')
ax.axis('on')

# choose a different color for each trajectory
colors = plt.cm.jet(np.linspace(0, 1, N_trajectories))
# set up trajectory lines
lines = sum([ax.plot([], [], [], '-', c=c) for c in colors], [])
# set up points
pts = sum([ax.plot([], [], [], 'o', c=c) for c in colors], [])
# set up lines which create the stick figures
stick_lines = sum([ax.plot([], [], [], '-', c=c) for c in colors], [])

# prepare the axes limits
ax.set_xlim(df_minmax.ix['x'].values)
ax.set_ylim(df_minmax.ix['z'].values) # note usage of z coordinate
ax.set_zlim(df_minmax.ix['y'].values) # note usage of y coordinate

# set point-of-view: specified by (altitude degrees, azimuth degrees)
ax.view_init(30, 0)

# initialization function: plot the background of each frame
def init():
    for line, pt, stick_line in zip(lines, pts, stick_lines):
        # trajectory lines
        line.set_data([], [])
        line.set_3d_properties([])
        # points
        pt.set_data([], [])
        pt.set_3d_properties([])
        # stick lines
        stick_line.set_data([], [])
        stick_line.set_3d_properties([])
    return lines + pts + stick_lines

# animation function.  This will be called sequentially with the frame number
def animate(i):
    # we'll step two time-steps per frame.  This leads to nice results.
    i = (5 * i) % x_t.shape[1]

    for line, pt, stick_line, xi in zip(lines, pts, stick_lines, x_t):
        x, z, y = xi[:i].T # note ordering of points to line up with true exogenous registration (x,z,y)
        # trajectory lines
        line.set_data(x,y)
        line.set_3d_properties(z)
        # points
        pt.set_data(x[-1:], y[-1:])
        pt.set_3d_properties(z[-1:])
        # stick lines
        #stick_line.set_data(xx,zz)
        #stick_line.set_3d_properties(yy)
    ax.view_init(30, 0.3 * i)
    fig.canvas.draw()
    return lines + pts + stick_lines

# instantiate the animator.
anim = animation.FuncAnimation(fig, animate, init_func=init, frames=500, interval=30, blit=True)

# Save as mp4. This requires mplayer or ffmpeg to be installed
#anim.save('lorentz_attractor.mp4', fps=15, extra_args=['-vcodec', 'libx264'])

plt.show()

So, to conclude: I would like lines that moves with the point pairs (orange, yellow) and (yellow, green). If someone could show me how to do that I should be able to extrapolate the methods to the rest of the animation.

As ever, any help is much appreciated.

The original data can be found here, if anyone wants to replicate: https://www.dropbox.com/sh/80f8ue4ffa4067t/Pntl5-gUW4

EDIT: IMPLEMENTED SOLUTION

Here is the final result, using the proposed solution.

Answer 1

I modified your code to add stick lines, but to simplify the code, I removed the trace lines:

import numpy as np
import pandas as pd
import matplotlib
matplotlib.use('TkAgg') # Need to use in order to run on mac
from matplotlib import pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from matplotlib.colors import cnames
from matplotlib import animation

#=============================================================================================

t_start = 1917 # start frame
t_end = 2130 # end frame

data = pd.read_csv('Smart-first_phase_NaN-zeros.csv') # only coordinate data
df = data.loc[t_start:t_end,'Shoulder_left_x':'Ankle_right_z']

# Find max and min values for animation ranges
df_minmax = pd.DataFrame(index=list('xyz'),columns=range(2))
for i in list('xyz'):
    c_max = df.filter(regex='_{}'.format(i)).max().max()
    c_min = df.filter(regex='_{}'.format(i)).min().min()
    df_minmax.ix[i] = np.array([c_min,c_max])

df_minmax = 1.3*df_minmax # increase by 30% to make animation look better

df.columns  = np.repeat(range(12),3) # store cols like this for simplicity
N_tag = df.shape[1]/3 # nr of tags used (all)

N_trajectories = N_tag

t = np.linspace(0,data.Time[t_end],df.shape[0]) # pseudo time-vector for first walking activity
x_t = np.zeros(shape=(N_tag,df.shape[0],3)) # empty animation array (3D)

for tag in range(12):
    # store data in numpy 3D array: (tag,time-stamp,xyz-coordinates)
    x_t[tag,:,:] = df[tag]

x_t = x_t[:, :, [0, 2, 1]]

# Set up figure & 3D axis for animation
fig = plt.figure()
ax = fig.add_axes([0, 0, 1, 1], projection='3d')
ax.axis('on')

# choose a different color for each trajectory
colors = plt.cm.jet(np.linspace(0, 1, N_trajectories))
# set up trajectory lines
lines = sum([ax.plot([], [], [], '-', c=c) for c in colors], [])
# set up points
pts = sum([ax.plot([], [], [], 'o', c=c) for c in colors], [])
# set up lines which create the stick figures

stick_defines = [
    (0, 1),
    (1, 2),
    (3, 4),
    (4, 5),
    (6, 7),
    (7, 8),
    (9, 10),
    (10, 11)
]

stick_lines = [ax.plot([], [], [], 'k-')[0] for _ in stick_defines]

# prepare the axes limits
ax.set_xlim(df_minmax.ix['x'].values)
ax.set_ylim(df_minmax.ix['z'].values) # note usage of z coordinate
ax.set_zlim(df_minmax.ix['y'].values) # note usage of y coordinate

# set point-of-view: specified by (altitude degrees, azimuth degrees)
ax.view_init(30, 0)

# initialization function: plot the background of each frame
def init():
    for line, pt in zip(lines, pts):
        # trajectory lines
        line.set_data([], [])
        line.set_3d_properties([])
        # points
        pt.set_data([], [])
        pt.set_3d_properties([])
    return lines + pts + stick_lines

# animation function.  This will be called sequentially with the frame number
def animate(i):
    # we'll step two time-steps per frame.  This leads to nice results.
    i = (5 * i) % x_t.shape[1]

    for line, pt, xi in zip(lines, pts, x_t):
        x, y, z = xi[:i].T # note ordering of points to line up with true exogenous registration (x,z,y)
        pt.set_data(x[-1:], y[-1:])
        pt.set_3d_properties(z[-1:])

    for stick_line, (sp, ep) in zip(stick_lines, stick_defines):
        stick_line._verts3d = x_t[[sp,ep], i, :].T.tolist()

    ax.view_init(30, 0.3 * i)
    fig.canvas.draw()
    return lines + pts + stick_lines

# instantiate the animator.
anim = animation.FuncAnimation(fig, animate, init_func=init, frames=500, interval=30, blit=True)

plt.show()

Here is one frame of the animation: