Python包安装——mayavi安装

深圳爱智慧科技有限公司 深度学习 人工智能 智能投顾 机器学习 中国版kensho 深圳爱智慧科技有限公司

安装mayavi有两种方法：

一种是安装集成了多个python包的软件，如 Enthought Canopy, Pythonxy, orAnaconda.可是貌似几个集成环境中都没有看到mayavi的影子。

第二种就是从源代码开始安装：安装mayavi以前先要安装这几个依赖包

You must have the following libraries installed before installing the Mayavi project:

• Numpy version 1.1.1 or later
• VTK version 5.0 or later
• wxPython version 2.8 or later
• configobj

那么先开始安装vtk吧。两种方法：

一：

首先安装vtk：

安装VTK，从网站http://www.vtk.org/VTK/resources/software.html 下载最新版本，此文安装的是：vtkpython-6.1.0-Windows-64bit.exe

3）配置VTK环境变量：

A：把VTK中的bin文件加入系统变量中：如：PATH：C:\Program Files\VTK 6.1.0\bin

B：设置python路径，在系统变量中建立PYTHONPATH ：如：PYTHONPATH ：C:\Program Files\VTK 6.1.0\bin;C:\Program Files\VTK 6.1.0\bin\Lib\site-packages

4）测试，打开python IDEL界面，输入：import vtk

>>> import vtk
>>> 

没报错，就说明安装成功了。

我是直接把site-packages里面的vtk拷贝到了python27下面的site-packages中

安装VTK的第二种方法，从源代码开始安装（须要先编译VTK，因为是win10系统没成功）

好吧，装了了vtk以后再到mayavi-4.4.3文件夹下运行python setup.py install 仍是不成功。

因而又老老实实看了Enthought Canopy发现vtk，tvtk还有traits啥的仍是有的，尽管就是没有mayavi，先装个Canopy，而后把把site-packages里面的vtk拷贝到了python27下面的site-packages中

我擦，不科学，怎么会这样

特么居然有了，有了！！！！！！直接考到python27下面的site-packages中

import mayavi.mlab as mlab

遇到ValueError: cannot set toolkit to wx because it has already been set to qt4的问题

从这个网址看到的解决方案

Hi, 

Building traitsui GUIs from the ipython notebook should work the same as from a regular ipython console. The difference between configure_traits and edit_traits is that the first one creates a GUI event loop before building the traitsUI window in it. That's what makes it responsive or not. Alternatively you can manually start the GUI event loop by specifying --gui=?? when you create ipython, for example:
ipython notebook --gui=wx
Then you should use edit_traits() and the UI element will be responsive. 

Finally, ETSConfig allows you to set the backend at your script runtime instead of when you start python/ipython. This is what I do the most so that I don't need to set the value of the ETSConfig environment variable, or set the gui switch when starting python/ipython.

To conclude, the following code is what I do the most and works from within the ipython notebook started with ipython notebook (version 0.13):
from traits.etsconfig.api import ETSConfig
ETSConfig.toolkit = 'qt4'

from traits.api import HasTraits, Enum
from traitsui.api import View

class A(HasTraits):
    x = Enum(["foo","bar"])

a = A()
a.configure_traits()

HTH, 
Jonathan

Thanks, Jonathan.

I would like to follow your advice, but am getting some errors. When I try this at the top of a new notebook, I get:

from traits.api import *

from traits.etsconfig.api import ETSConfig

from traitsui.api import *

ETSConfig.toolkit = 'qt4'

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-2-8a5daaa83ef1> in <module>()
----> 1 ETSConfig.toolkit = 'qt4'

/Library/Frameworks/Python.framework/Versions/7.3/lib/python2.7/site-packages/traits/etsconfig/etsconfig.py in _set_toolkit(self, toolkit)
 208         if self._toolkit and self._toolkit != toolkit:
 209             raise ValueError, "cannot set toolkit to %s because it has "\
--> 210 "already been set to %s" % (toolkit, self._toolkit)
 211 
 212         self._toolkit = toolkit

ValueError: cannot set toolkit to qt4 because it has already been set to wx

Any idea what might be causing this? I suspect the other error below on starting ipython-qtconsole may be related, but am equally clueless about this. I think both these worked until recently. 

Ah, Google to the rescue

https://mail.enthought.com/pipermail/enthought-dev/2009-August/023406.html

I must put ETSConfig.toolkit = 'qt4' before any other ETS import, or the earlier import will set default to wx. 

The corresponding ipython-qtconsole error happened because I still had an ipython notebook server open in which the same error had already happened, seems OK now.

Why one uses 2.6 and the other 2.7 I don't even need to know, as long as it works :-)

All set now, thank you!

================================================================================================================================

下面是一个例子

import numpy as np

import mayavi . mlab as mlab

import    moviepy . editor as mpy

 

duration = 2 # duration of the animation in seconds (it will loop)

 

# MAKE A FIGURE WITH MAYAVI

 

fig = mlab . figure ( size = ( 500 , 500 ) , bgcolor = ( 1 , 1 , 1 ) )

 

u = np . linspace ( 0 , 2 * np . pi , 100 )

xx , yy , zz = np . cos ( u ) , np . sin ( 3 * u ) , np . sin ( u ) # Points

l = mlab . plot3d ( xx , yy , zz , representation = "wireframe" , tube_sides = 5 ,

                 line_width = . 5 , tube_radius = 0.2 , figure = fig )

 

# ANIMATE THE FIGURE WITH MOVIEPY, WRITE AN ANIMATED GIF

 

def make_frame ( t ) :

     """ Generates and returns the frame for time t. """

     y = np . sin ( 3 * u ) * ( 0.2 + 0.5 * np . cos ( 2 * np . pi * t / duration ) )

     l . mlab_source . set ( y = y ) # change y-coordinates of the mesh

     mlab . view ( azimuth = 360 * t / duration , distance = 9 ) # camera angle

     return mlab . screenshot ( antialiased = True ) # return a RGB image

 

animation = mpy . VideoClip ( make_frame , duration = duration ) . resize ( 0.5 )

# Video generation takes 10 seconds, GIF generation takes 25s

animation . write_videofile ( "wireframe.mp4" , fps = 20 )

animation . write_gif ( "wireframe.gif" , fps = 20 )

运行这个例子提醒没有安装“ffmpeg.win32.exe”

从https://github.com/imageio/imageio-binaries/blob/master/ffmpeg/ffmpeg.win32.exe下载了个“ffmpeg.win32.exe”文件。

而后在用户和系统环境变量里面新建一个名字叫“IMAGEIO_FFMPEG_EXE”，值是你存放这个exe文件的地址的变量。