Re: [Matplotlib-users] Plotting style
Le 06/03/2015 22:42, Marin GILLES a écrit : This package is indeeed pretty nice, and I will surely take a look into it, but the way styles are added does not seem quite practical or shareable. In my opinion, having a style file for each paper makes things more flexible, although this package may get more control out of the box. Also, not being built-in makes you install an other package, and I think some people either do not want to do it, nor know how to do it. On an other topic, I started working on some of the features you wanted to integrate with your PR https://github.com/matplotlib/matplotlib/pull/2702. I guessed that when you talked about adding the |set_ticks_location| to the rcParams, you wanted to control whether the ticks are in or out of the axes box? Finally, I added a |style| parameter to the rcParams. It lets you choose from your |matplotlibrc| which style you want to use. On top of that, I made it recursive, so that you can design a style directly from other styles. The only thing I could not get to work was to have your style loading directly when importing matplotlib (when defining from your rc file). You actually have to import the |matplotlib.style| lib to get your rc defined style to load up. I will continue working on the other features described in olga’s PR https://github.com/matplotlib/matplotlib/pull/2702 before submitting one on my own. But if you want to take a look, and tell me how I can improve what I did, you can find it on my repo https://github.com/Mrngilles/matplotlib. Thanks Marin Le 06/03/2015 22:18, Olga Botvinnik a écrit : There's also the plotsettings package which makes it easy to switch between styles required by different papers. https://pypi.python.org/pypi/plotsettings On Wed, Mar 4, 2015 at 1:29 PM Marin GILLES mrngil...@gmail.com mailto:mrngil...@gmail.com wrote: Le 04/03/2015 06:21, Tony Yu a écrit : On Tue, Mar 3, 2015 at 11:50 AM, Gökhan Sever gokhanse...@gmail.com mailto:gokhanse...@gmail.com wrote: I see seaborn has paper, notebook, talk, and poster options. http://stanford.edu/~mwaskom/software/seaborn-dev/aesthetics.html http://stanford.edu/%7Emwaskom/software/seaborn-dev/aesthetics.html Apperantly he scales each parameter to get modified views. This would be a good addition for any of the styles available in matplotlib. A similar pattern with `matplotlib.style` would use chained stylesheets. The idea would be to make stylesheets either aesthetics focused or layout focused. By aesthetics, I mean things like colors and marker shape, and by layout, I mean things like default figure size, figure padding, font size, etc. Then you can easily have a style that defines the general aesthetics and easily modify it for papers, talks, etc. Here's an example from `mpltools`, but the same syntax applies to the `style` module in `matplotlib`: http://tonysyu.github.io/mpltools/auto_examples/style/plot_multiple_styles.html (PoF = Physics of Fluids journal; IIRC I think I have some personal stylesheets that take the normal two-column figure layout and convert it to a full-page layout.) -Tony On Tue, Mar 3, 2015 at 12:35 PM, Marin GILLES mrngil...@gmail.com mailto:mrngil...@gmail.com wrote: Le 03/03/2015 18:15, Gökhan Sever a écrit : On Tue, Mar 3, 2015 at 12:07 PM, Marin GILLES mrngil...@gmail.com mailto:mrngil...@gmail.com wrote: Sure, I'll be careful about that. I'm going to go try and design some new interesting ones. Maybe adding some styles specific to some plot types could be useful. Also some styles specific for some applications (geoscience, biology)? If you have any other ideas, please let me know. -- *Marin GILLES* It would be good to have styles for paper and presentation modes. The former would have smaller ticks, labels, linewidths, other axis elements that goes into a journal publication, while the latter with much magnified elements to be clearly visible on a screen from the back of a room. Indeed it would be a very good idea. I've seen that already in the seaborn lib I guess. -- *Marin GILLES* /PhD student CNRS / /Laboratoire Interdisciplinaire Carnot de Bourgogne (ICB) UMR 6303 CNRS - Université de Bourgogne 9 av Alain Savary, BP 47870 21078, Dijon (France) / ☎ (+33)6.79.35.30.11 tel:%28%2B33%296.79.35.30.11 ✉ marin.gil...@u-bourgogne.fr mailto:marin.gil...@u-bourgogne.fr -- Gökhan
Re: [Matplotlib-users] Lorenz - solution
I think you need to ask Jake Vanderplas -- the code
is all his! See the link in the email to get to his blog...
Thanks again!
On Tue, Mar 10, 2015 at 8:49 AM, Benjamin Root ben.r...@ou.edu wrote:
+1000!!
Great job! Would you mind if I clean it up a bit and add it to the
mplot3d/animation gallery? Full credit, of course.
On Tue, Mar 10, 2015 at 11:30 AM, Prahas David Nafissian
prahas.mu...@gmail.com wrote:
Friends,
I thought you'd like to see the solution.
Many thanks to Jake Vanderplas for his code and teachings:
https://jakevdp.github.io/blog/2013/02/16/animating-the-lorentz-system-in-3d/
If you start a new IP Notebook session, run as your first entry:
%pylab
and then copy and paste the text below and run it, you should be good to
go
(on a Mac, at least).
There are several parameters I've changed from his original, and I've
commented as I've changed. The original code is at the link above.
There is one error in his code -- I've documented it below.
Again, thanks to the community, Jake, and Ben Root.
--Prahas
**
import numpy as np
from scipy import integrate
from matplotlib import pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from matplotlib.colors import cnames
from matplotlib import animation
# orig value of N_traj was 20 -- very cool this way.
N_trajectories = 1
def lorentz_deriv((x, y, z), t0, sigma=10., beta=8./3, rho=28.0):
Compute the time-derivative of a Lorentz system.
return [sigma * (y - x), x * (rho - z) - y, x * y - beta * z]
# Choose random starting points, uniformly distributed from -15 to 15
np.random.seed(1)
# changing from -15,30 to 10,5 below starts the drawing in the middle,
# rather than getting the long line from below
# if using N_Traj 1, return to orig values.
# x0 = -15 + 30 * np.random.random((N_trajectories, 3))
x0 = 10 + 5 * np.random.random((N_trajectories, 3))
# Solve for the trajectories
# orig values: 0,4,1000
# 3rd value -- lower it, it gets choppier.
# 2nd value -- increase it -- more points, but speedier.
# change middle num from 4 to 15 -- this adds points
t = np.linspace(0, 40, 3000)
x_t = np.asarray([integrate.odeint(lorentz_deriv, x0i, t)
for x0i in x0])
# Set up figure 3D axis for animation
fig = plt.figure()
ax = fig.add_axes([0, 0, 1, 1], projection='3d')
# changing off to on below adds axises. slows it down but you
# can fix that with interval value in the animation call
ax.axis('on')
# choose a different color for each trajectory
colors = plt.cm.jet(np.linspace(0, 1, N_trajectories))
# set up lines and points -- this is a correction from
# the orig jake code. the next four lines...
lines = [ax.plot([], [], [], '-', c=c)[0]
for c in colors]
pts = [ax.plot([], [], [], 'o', c=c)[0]
for c in colors]
# prepare the axes limits
ax.set_xlim((-25, 25))
ax.set_ylim((-35, 35))
ax.set_zlim((5, 55))
# 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):
line.set_data([], [])
line.set_3d_properties([])
pt.set_data([], [])
pt.set_3d_properties([])
return lines + pts
# 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 = (2 * i) % x_t.shape[1]
for line, pt, xi in zip(lines, pts, x_t):
x, y, z = xi[:i].T
line.set_data(x, y)
line.set_3d_properties(z)
pt.set_data(x[-1:], y[-1:])
pt.set_3d_properties(z[-1:])
# changed 0.3 to 0.05 below -- this slows the rotation of the
view.
# changed 30 to 20 below
# changing 20 to (20 + (.1 * i)) rotates on the Z axis. trippy.
ax.view_init(10, 0.1 * i)
# ax.view_init(10, 100)
fig.canvas.draw()
return lines + pts
# instantiate the animator. I've deleted the blit switch (for Mac)
# enlarging frames=500 works now -- it failed before because I didn't give
it
# enough data -- by changing the t=np.linspace line above I generate
more points.
# interval larger slows it down
# changed inteval from 30 to 200, frames from 500 to 3000
anim = animation.FuncAnimation(fig, animate, init_func=init,
frames=3000, interval=200)
# Save as mp4. This requires mplayer or ffmpeg to be installed.
COMPLEX!
# Instead, use a screen record program: Quicktime on the Mac; MS
Expression Encoder on PC.
# anim.save('PDNlorentz_attractor.mp4', fps=15, extra_args=['-vcodec',
'libx264'])
plt.show()
--
Dive into the World of Parallel Programming The Go Parallel Website,
sponsored
by Intel and developed in partnership with
Re: [Matplotlib-users] Lorenz - solution
@ Adam -- thanks!
@ Everyone: given the Lorenz code shared, is there a way
to generate a log file of the x,y,z points generated?
Thanks in advance.
--Prahas
On Tue, Mar 10, 2015 at 8:59 AM, Adam Hughes hughesada...@gmail.com wrote:
That's pretty swag!
On Tue, Mar 10, 2015 at 11:49 AM, Benjamin Root ben.r...@ou.edu wrote:
+1000!!
Great job! Would you mind if I clean it up a bit and add it to the
mplot3d/animation gallery? Full credit, of course.
On Tue, Mar 10, 2015 at 11:30 AM, Prahas David Nafissian
prahas.mu...@gmail.com wrote:
Friends,
I thought you'd like to see the solution.
Many thanks to Jake Vanderplas for his code and teachings:
https://jakevdp.github.io/blog/2013/02/16/animating-the-lorentz-system-in-3d/
If you start a new IP Notebook session, run as your first entry:
%pylab
and then copy and paste the text below and run it, you should be good to
go
(on a Mac, at least).
There are several parameters I've changed from his original, and I've
commented as I've changed. The original code is at the link above.
There is one error in his code -- I've documented it below.
Again, thanks to the community, Jake, and Ben Root.
--Prahas
**
import numpy as np
from scipy import integrate
from matplotlib import pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from matplotlib.colors import cnames
from matplotlib import animation
# orig value of N_traj was 20 -- very cool this way.
N_trajectories = 1
def lorentz_deriv((x, y, z), t0, sigma=10., beta=8./3, rho=28.0):
Compute the time-derivative of a Lorentz system.
return [sigma * (y - x), x * (rho - z) - y, x * y - beta * z]
# Choose random starting points, uniformly distributed from -15 to 15
np.random.seed(1)
# changing from -15,30 to 10,5 below starts the drawing in the middle,
# rather than getting the long line from below
# if using N_Traj 1, return to orig values.
# x0 = -15 + 30 * np.random.random((N_trajectories, 3))
x0 = 10 + 5 * np.random.random((N_trajectories, 3))
# Solve for the trajectories
# orig values: 0,4,1000
# 3rd value -- lower it, it gets choppier.
# 2nd value -- increase it -- more points, but speedier.
# change middle num from 4 to 15 -- this adds points
t = np.linspace(0, 40, 3000)
x_t = np.asarray([integrate.odeint(lorentz_deriv, x0i, t)
for x0i in x0])
# Set up figure 3D axis for animation
fig = plt.figure()
ax = fig.add_axes([0, 0, 1, 1], projection='3d')
# changing off to on below adds axises. slows it down but you
# can fix that with interval value in the animation call
ax.axis('on')
# choose a different color for each trajectory
colors = plt.cm.jet(np.linspace(0, 1, N_trajectories))
# set up lines and points -- this is a correction from
# the orig jake code. the next four lines...
lines = [ax.plot([], [], [], '-', c=c)[0]
for c in colors]
pts = [ax.plot([], [], [], 'o', c=c)[0]
for c in colors]
# prepare the axes limits
ax.set_xlim((-25, 25))
ax.set_ylim((-35, 35))
ax.set_zlim((5, 55))
# 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):
line.set_data([], [])
line.set_3d_properties([])
pt.set_data([], [])
pt.set_3d_properties([])
return lines + pts
# 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 = (2 * i) % x_t.shape[1]
for line, pt, xi in zip(lines, pts, x_t):
x, y, z = xi[:i].T
line.set_data(x, y)
line.set_3d_properties(z)
pt.set_data(x[-1:], y[-1:])
pt.set_3d_properties(z[-1:])
# changed 0.3 to 0.05 below -- this slows the rotation of the
view.
# changed 30 to 20 below
# changing 20 to (20 + (.1 * i)) rotates on the Z axis. trippy.
ax.view_init(10, 0.1 * i)
# ax.view_init(10, 100)
fig.canvas.draw()
return lines + pts
# instantiate the animator. I've deleted the blit switch (for Mac)
# enlarging frames=500 works now -- it failed before because I didn't
give it
# enough data -- by changing the t=np.linspace line above I generate
more points.
# interval larger slows it down
# changed inteval from 30 to 200, frames from 500 to 3000
anim = animation.FuncAnimation(fig, animate, init_func=init,
frames=3000, interval=200)
# Save as mp4. This requires mplayer or ffmpeg to be installed.
COMPLEX!
# Instead, use a screen record program: Quicktime on the Mac; MS
Expression Encoder on PC.
# anim.save('PDNlorentz_attractor.mp4', fps=15, extra_args=['-vcodec',
'libx264'])
plt.show()
Re: [Matplotlib-users] Lorenz - solution
+1000!!
Great job! Would you mind if I clean it up a bit and add it to the
mplot3d/animation gallery? Full credit, of course.
On Tue, Mar 10, 2015 at 11:30 AM, Prahas David Nafissian
prahas.mu...@gmail.com wrote:
Friends,
I thought you'd like to see the solution.
Many thanks to Jake Vanderplas for his code and teachings:
https://jakevdp.github.io/blog/2013/02/16/animating-the-lorentz-system-in-3d/
If you start a new IP Notebook session, run as your first entry:
%pylab
and then copy and paste the text below and run it, you should be good to go
(on a Mac, at least).
There are several parameters I've changed from his original, and I've
commented as I've changed. The original code is at the link above.
There is one error in his code -- I've documented it below.
Again, thanks to the community, Jake, and Ben Root.
--Prahas
**
import numpy as np
from scipy import integrate
from matplotlib import pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from matplotlib.colors import cnames
from matplotlib import animation
# orig value of N_traj was 20 -- very cool this way.
N_trajectories = 1
def lorentz_deriv((x, y, z), t0, sigma=10., beta=8./3, rho=28.0):
Compute the time-derivative of a Lorentz system.
return [sigma * (y - x), x * (rho - z) - y, x * y - beta * z]
# Choose random starting points, uniformly distributed from -15 to 15
np.random.seed(1)
# changing from -15,30 to 10,5 below starts the drawing in the middle,
# rather than getting the long line from below
# if using N_Traj 1, return to orig values.
# x0 = -15 + 30 * np.random.random((N_trajectories, 3))
x0 = 10 + 5 * np.random.random((N_trajectories, 3))
# Solve for the trajectories
# orig values: 0,4,1000
# 3rd value -- lower it, it gets choppier.
# 2nd value -- increase it -- more points, but speedier.
# change middle num from 4 to 15 -- this adds points
t = np.linspace(0, 40, 3000)
x_t = np.asarray([integrate.odeint(lorentz_deriv, x0i, t)
for x0i in x0])
# Set up figure 3D axis for animation
fig = plt.figure()
ax = fig.add_axes([0, 0, 1, 1], projection='3d')
# changing off to on below adds axises. slows it down but you
# can fix that with interval value in the animation call
ax.axis('on')
# choose a different color for each trajectory
colors = plt.cm.jet(np.linspace(0, 1, N_trajectories))
# set up lines and points -- this is a correction from
# the orig jake code. the next four lines...
lines = [ax.plot([], [], [], '-', c=c)[0]
for c in colors]
pts = [ax.plot([], [], [], 'o', c=c)[0]
for c in colors]
# prepare the axes limits
ax.set_xlim((-25, 25))
ax.set_ylim((-35, 35))
ax.set_zlim((5, 55))
# 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):
line.set_data([], [])
line.set_3d_properties([])
pt.set_data([], [])
pt.set_3d_properties([])
return lines + pts
# 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 = (2 * i) % x_t.shape[1]
for line, pt, xi in zip(lines, pts, x_t):
x, y, z = xi[:i].T
line.set_data(x, y)
line.set_3d_properties(z)
pt.set_data(x[-1:], y[-1:])
pt.set_3d_properties(z[-1:])
# changed 0.3 to 0.05 below -- this slows the rotation of the view.
# changed 30 to 20 below
# changing 20 to (20 + (.1 * i)) rotates on the Z axis. trippy.
ax.view_init(10, 0.1 * i)
# ax.view_init(10, 100)
fig.canvas.draw()
return lines + pts
# instantiate the animator. I've deleted the blit switch (for Mac)
# enlarging frames=500 works now -- it failed before because I didn't give
it
# enough data -- by changing the t=np.linspace line above I generate
more points.
# interval larger slows it down
# changed inteval from 30 to 200, frames from 500 to 3000
anim = animation.FuncAnimation(fig, animate, init_func=init,
frames=3000, interval=200)
# Save as mp4. This requires mplayer or ffmpeg to be installed.
COMPLEX!
# Instead, use a screen record program: Quicktime on the Mac; MS
Expression Encoder on PC.
# anim.save('PDNlorentz_attractor.mp4', fps=15, extra_args=['-vcodec',
'libx264'])
plt.show()
--
Dive into the World of Parallel Programming The Go Parallel Website,
sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for
all
things parallel software development, from weekly thought leadership blogs
to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now.
[Matplotlib-users] Lorenz - solution
Friends,
I thought you'd like to see the solution.
Many thanks to Jake Vanderplas for his code and teachings:
https://jakevdp.github.io/blog/2013/02/16/animating-the-lorentz-system-in-3d/
If you start a new IP Notebook session, run as your first entry:
%pylab
and then copy and paste the text below and run it, you should be good to go
(on a Mac, at least).
There are several parameters I've changed from his original, and I've
commented as I've changed. The original code is at the link above.
There is one error in his code -- I've documented it below.
Again, thanks to the community, Jake, and Ben Root.
--Prahas
**
import numpy as np
from scipy import integrate
from matplotlib import pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from matplotlib.colors import cnames
from matplotlib import animation
# orig value of N_traj was 20 -- very cool this way.
N_trajectories = 1
def lorentz_deriv((x, y, z), t0, sigma=10., beta=8./3, rho=28.0):
Compute the time-derivative of a Lorentz system.
return [sigma * (y - x), x * (rho - z) - y, x * y - beta * z]
# Choose random starting points, uniformly distributed from -15 to 15
np.random.seed(1)
# changing from -15,30 to 10,5 below starts the drawing in the middle,
# rather than getting the long line from below
# if using N_Traj 1, return to orig values.
# x0 = -15 + 30 * np.random.random((N_trajectories, 3))
x0 = 10 + 5 * np.random.random((N_trajectories, 3))
# Solve for the trajectories
# orig values: 0,4,1000
# 3rd value -- lower it, it gets choppier.
# 2nd value -- increase it -- more points, but speedier.
# change middle num from 4 to 15 -- this adds points
t = np.linspace(0, 40, 3000)
x_t = np.asarray([integrate.odeint(lorentz_deriv, x0i, t)
for x0i in x0])
# Set up figure 3D axis for animation
fig = plt.figure()
ax = fig.add_axes([0, 0, 1, 1], projection='3d')
# changing off to on below adds axises. slows it down but you
# can fix that with interval value in the animation call
ax.axis('on')
# choose a different color for each trajectory
colors = plt.cm.jet(np.linspace(0, 1, N_trajectories))
# set up lines and points -- this is a correction from
# the orig jake code. the next four lines...
lines = [ax.plot([], [], [], '-', c=c)[0]
for c in colors]
pts = [ax.plot([], [], [], 'o', c=c)[0]
for c in colors]
# prepare the axes limits
ax.set_xlim((-25, 25))
ax.set_ylim((-35, 35))
ax.set_zlim((5, 55))
# 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):
line.set_data([], [])
line.set_3d_properties([])
pt.set_data([], [])
pt.set_3d_properties([])
return lines + pts
# 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 = (2 * i) % x_t.shape[1]
for line, pt, xi in zip(lines, pts, x_t):
x, y, z = xi[:i].T
line.set_data(x, y)
line.set_3d_properties(z)
pt.set_data(x[-1:], y[-1:])
pt.set_3d_properties(z[-1:])
# changed 0.3 to 0.05 below -- this slows the rotation of the view.
# changed 30 to 20 below
# changing 20 to (20 + (.1 * i)) rotates on the Z axis. trippy.
ax.view_init(10, 0.1 * i)
# ax.view_init(10, 100)
fig.canvas.draw()
return lines + pts
# instantiate the animator. I've deleted the blit switch (for Mac)
# enlarging frames=500 works now -- it failed before because I didn't give it
# enough data -- by changing the t=np.linspace line above I generate
more points.
# interval larger slows it down
# changed inteval from 30 to 200, frames from 500 to 3000
anim = animation.FuncAnimation(fig, animate, init_func=init,
frames=3000, interval=200)
# Save as mp4. This requires mplayer or ffmpeg to be installed. COMPLEX!
# Instead, use a screen record program: Quicktime on the Mac; MS
Expression Encoder on PC.
# anim.save('PDNlorentz_attractor.mp4', fps=15, extra_args=['-vcodec',
'libx264'])
plt.show()
--
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now. http://goparallel.sourceforge.net/
___
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] Lorenz - solution
That's pretty swag!
On Tue, Mar 10, 2015 at 11:49 AM, Benjamin Root ben.r...@ou.edu wrote:
+1000!!
Great job! Would you mind if I clean it up a bit and add it to the
mplot3d/animation gallery? Full credit, of course.
On Tue, Mar 10, 2015 at 11:30 AM, Prahas David Nafissian
prahas.mu...@gmail.com wrote:
Friends,
I thought you'd like to see the solution.
Many thanks to Jake Vanderplas for his code and teachings:
https://jakevdp.github.io/blog/2013/02/16/animating-the-lorentz-system-in-3d/
If you start a new IP Notebook session, run as your first entry:
%pylab
and then copy and paste the text below and run it, you should be good to
go
(on a Mac, at least).
There are several parameters I've changed from his original, and I've
commented as I've changed. The original code is at the link above.
There is one error in his code -- I've documented it below.
Again, thanks to the community, Jake, and Ben Root.
--Prahas
**
import numpy as np
from scipy import integrate
from matplotlib import pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from matplotlib.colors import cnames
from matplotlib import animation
# orig value of N_traj was 20 -- very cool this way.
N_trajectories = 1
def lorentz_deriv((x, y, z), t0, sigma=10., beta=8./3, rho=28.0):
Compute the time-derivative of a Lorentz system.
return [sigma * (y - x), x * (rho - z) - y, x * y - beta * z]
# Choose random starting points, uniformly distributed from -15 to 15
np.random.seed(1)
# changing from -15,30 to 10,5 below starts the drawing in the middle,
# rather than getting the long line from below
# if using N_Traj 1, return to orig values.
# x0 = -15 + 30 * np.random.random((N_trajectories, 3))
x0 = 10 + 5 * np.random.random((N_trajectories, 3))
# Solve for the trajectories
# orig values: 0,4,1000
# 3rd value -- lower it, it gets choppier.
# 2nd value -- increase it -- more points, but speedier.
# change middle num from 4 to 15 -- this adds points
t = np.linspace(0, 40, 3000)
x_t = np.asarray([integrate.odeint(lorentz_deriv, x0i, t)
for x0i in x0])
# Set up figure 3D axis for animation
fig = plt.figure()
ax = fig.add_axes([0, 0, 1, 1], projection='3d')
# changing off to on below adds axises. slows it down but you
# can fix that with interval value in the animation call
ax.axis('on')
# choose a different color for each trajectory
colors = plt.cm.jet(np.linspace(0, 1, N_trajectories))
# set up lines and points -- this is a correction from
# the orig jake code. the next four lines...
lines = [ax.plot([], [], [], '-', c=c)[0]
for c in colors]
pts = [ax.plot([], [], [], 'o', c=c)[0]
for c in colors]
# prepare the axes limits
ax.set_xlim((-25, 25))
ax.set_ylim((-35, 35))
ax.set_zlim((5, 55))
# 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):
line.set_data([], [])
line.set_3d_properties([])
pt.set_data([], [])
pt.set_3d_properties([])
return lines + pts
# 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 = (2 * i) % x_t.shape[1]
for line, pt, xi in zip(lines, pts, x_t):
x, y, z = xi[:i].T
line.set_data(x, y)
line.set_3d_properties(z)
pt.set_data(x[-1:], y[-1:])
pt.set_3d_properties(z[-1:])
# changed 0.3 to 0.05 below -- this slows the rotation of the
view.
# changed 30 to 20 below
# changing 20 to (20 + (.1 * i)) rotates on the Z axis. trippy.
ax.view_init(10, 0.1 * i)
# ax.view_init(10, 100)
fig.canvas.draw()
return lines + pts
# instantiate the animator. I've deleted the blit switch (for Mac)
# enlarging frames=500 works now -- it failed before because I didn't
give it
# enough data -- by changing the t=np.linspace line above I generate
more points.
# interval larger slows it down
# changed inteval from 30 to 200, frames from 500 to 3000
anim = animation.FuncAnimation(fig, animate, init_func=init,
frames=3000, interval=200)
# Save as mp4. This requires mplayer or ffmpeg to be installed.
COMPLEX!
# Instead, use a screen record program: Quicktime on the Mac; MS
Expression Encoder on PC.
# anim.save('PDNlorentz_attractor.mp4', fps=15, extra_args=['-vcodec',
'libx264'])
plt.show()
--
Dive into the World of Parallel Programming The Go Parallel Website,
sponsored
by Intel and developed in partnership with Slashdot Media, is your hub
for all
things parallel software development, from weekly thought leadership
Re: [Matplotlib-users] Plotting style
On 2015/03/09 8:14 PM, Marin GILLES wrote: Hi, As suggested in PR 2702 https://github.com/matplotlib/matplotlib/pull/2702, I have been trying to tell |scatter| to |get_current_color_cycle| for the facecolor. I guess I can use |axes.get_color()|to get the current color in the color cycle. However, I was not able to try this, as when I try to import pyplot I get an |ImportError: No module named _path|. It seems to be library related, but I’m not quite sure how I can solve this… It sounds like your installation is broken; _path is an extension module compiled from C++, and central to matplotlib's functionality. In what environment are you working? Did this failure arise after you modified code and then executed python setup.py install or something of that sort? Eric -- Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ ___ Matplotlib-users mailing list Matplotlib-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] Plotting style
Le 10/03/2015 07:52, Eric Firing a écrit : On 2015/03/09 8:14 PM, Marin GILLES wrote: Hi, As suggested in PR 2702 https://github.com/matplotlib/matplotlib/pull/2702, I have been trying to tell |scatter| to |get_current_color_cycle| for the facecolor. I guess I can use |axes.get_color()|to get the current color in the color cycle. However, I was not able to try this, as when I try to import pyplot I get an |ImportError: No module named _path|. It seems to be library related, but I’m not quite sure how I can solve this… It sounds like your installation is broken; _path is an extension module compiled from C++, and central to matplotlib's functionality. In what environment are you working? Did this failure arise after you modified code and then executed python setup.py install or something of that sort? Eric -- Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ ___ Matplotlib-users mailing list Matplotlib-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/matplotlib-users Actually, I just brute loaded mpl for source... I am not really used to it. So I guess I'll have to make a virtual env and install mpl in it? -- *Marin GILLES* /PhD student CNRS / /Laboratoire Interdisciplinaire Carnot de Bourgogne (ICB) UMR 6303 CNRS - Université de Bourgogne 9 av Alain Savary, BP 47870 21078, Dijon (France) / ☎ (+33)6.79.35.30.11 ✉ marin.gil...@u-bourgogne.fr mailto:marin.gil...@u-bourgogne.fr -- Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/___ Matplotlib-users mailing list Matplotlib-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/matplotlib-users
Re: [Matplotlib-users] Plotting style
On 2015/03/09 8:56 PM, Marin GILLES wrote: Actually, I just brute loaded mpl for source... I am not really used to it. So I guess I'll have to make a virtual env and install mpl in it? You have to build and install it somewhere, where it will be found when you try to import it; whether you use a virtual env is up to you. I managed for years without using virtual envs. Recently I've found them quite helpful, but a bit tricky and confusing at times. Eric -- Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ ___ Matplotlib-users mailing list Matplotlib-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/matplotlib-users
