我正在尝试使用python中的pcolor绘制DFT(离散傅立叶变换)图。我以前一直在使用Mathematica 8.0来实现这一点，但我发现mathematica8.0中的colorbar与我试图表示的数据有着糟糕的一一对应关系。例如，下面是我正在绘制的数据：[[0.,0.,0.10664,0.,0.,0.,0.0412719,0.,0.,0.],

[0.,0.351894,0.,0.17873,0.,0.,0.,0.,0.,0.],

[0.10663,0.,0.178183,0.,0.,0.,0.0405148,0.,0.,0.],

[0.,0.177586,0.,0.,0.,0.0500377,0.,0.,0.,0.],

[0.,0.,0.,0.,0.0588906,0.,0.,0.,0.,0.],

[0.,0.,0.,0.0493811,0.,0.,0.,0.,0.,0.],

[0.0397341,0.,0.0399249,0.,0.,0.,0.,0.,0.,0.],

[0.,0.,0.,0.,0.,0.,0.,0.,0.,0.],

[0.,0.,0.,0.,0.,0.,0.,0.,0.,0.],

[0.,0.,0.,0.,0.,0.,0.,0.,0.,0.]]

所以，它在DFT矩阵中有很多零或小数目，或者有少量的高频能量。

当我使用mathematica绘制此图时，结果如下：

颜色栏关闭了，我想我想用python来代替。

我的python代码(我从here中劫持的代码)是：from numpy import corrcoef, sum, log, arange

from numpy.random import rand

#from pylab import pcolor, show, colorbar, xticks, yticks

from pylab import *

data = np.array([[0.,0.,0.10664,0.,0.,0.,0.0412719,0.,0.,0.],

[0.,0.351894,0.,0.17873,0.,0.,0.,0.,0.,0.],

[0.10663,0.,0.178183,0.,0.,0.,0.0405148,0.,0.,0.],

[0.,0.177586,0.,0.,0.,0.0500377,0.,0.,0.,0.],

[0.,0.,0.,0.,0.0588906,0.,0.,0.,0.,0.],

[0.,0.,0.,0.0493811,0.,0.,0.,0.,0.,0.],

[0.0397341,0.,0.0399249,0.,0.,0.,0.,0.,0.,0.],

[0.,0.,0.,0.,0.,0.,0.,0.,0.,0.],

[0.,0.,0.,0.,0.,0.,0.,0.,0.,0.],

[0.,0.,0.,0.,0.,0.,0.,0.,0.,0.]], np.float)

pcolor(data)

colorbar()

yticks(arange(0.5,10.5),range(0,10))

xticks(arange(0.5,10.5),range(0,10))

#show()

savefig('/home/mydir/foo.eps',figsize=(4,4),dpi=100)

这段python代码绘制为：

下面是我的问题/问题清单：

我喜欢python是如何设计的，我想用这个但是。。。我怎样才能使所有代表“0”的“蓝色”像在我的mathematica图中那样消失呢？

如何旋转绘图以在左上角显示亮红色点？

我设置“dpi”的方式对吗？

有什么有用的参考资料可以用来加强我对python的喜爱吗？

我已经浏览了这里的其他问题和numpy的用户手册，但没有找到太多帮助。

我计划发布这些数据，这是相当重要的，我得到所有的位和块正确！:)

编辑：

修改的python代码和结果图！有人会对此提出什么改进建议，使其具有出版价值？from numpy import corrcoef, sum, log, arange, save

from numpy.random import rand

from pylab import *

data = np.array([[0.,0.,0.10664,0.,0.,0.,0.0412719,0.,0.,0.],

[0.,0.351894,0.,0.17873,0.,0.,0.,0.,0.,0.],

[0.10663,0.,0.178183,0.,0.,0.,0.0405148,0.,0.,0.],

[0.,0.177586,0.,0.,0.,0.0500377,0.,0.,0.,0.],

[0.,0.,0.,0.,0.0588906,0.,0.,0.,0.,0.],

[0.,0.,0.,0.0493811,0.,0.,0.,0.,0.,0.],

[0.0397341,0.,0.0399249,0.,0.,0.,0.,0.,0.,0.],

[0.,0.,0.,0.,0.,0.,0.,0.,0.,0.],

[0.,0.,0.,0.,0.,0.,0.,0.,0.,0.],

[0.,0.,0.,0.,0.,0.,0.,0.,0.,0.]], np.float)

v1 = abs(data).max()

v2 = abs(data).min()

pcolor(data, cmap="binary")

colorbar()

#xlabel("X", fontsize=12, fontweight="bold")

#ylabel("Y", fontsize=12, fontweight="bold")

xticks(arange(0.5,10.5),range(0,10),fontsize=19)

yticks(arange(0.5,10.5),range(0,10),fontsize=19)

axis([0,7,0,7])

#show()

savefig('/home/mydir/Desktop/py_dft.eps',figsize=(4,4),dpi=600)

画一个网络节点连接关系色彩图代码

主要调用函数pcolor, colorbar, colormap

clear;clc;

map=2*(rand(10)-0.5);

pcolor(map);

colorbar;

axis off;

其中colorbar函数会自动匹配map里面值的大小。

结果如下：

但是细看会有问题，画的图少了一行，查看pcolor帮助文档

里面说默认的shadding模式为faceted，shadding总共3种模式，faceted(默认的)，interp(对每个像素点插值)，flat(去除方块黑线)。

还有就是pcolor如果不设定X，Y的话是按照1—2，2—3，3—4……作为一块涂色，为此设置X,Y，让画图小方块移到0.5—1.5，1.5—2.5……，这样一会儿可以直接在小方块对应的中心设置坐标。

function result=plotMatricNet_15(Features,Featuresweight)

figure

% 由于pcolor函数不处理最后一行和最后一列，所以这里补一行一列

weightMetrics=zeros(16);

Featuresweight=abs(Featuresweight);

for i=1:length(Features)

edge=Features(i);

[row,col]=electrodesIndex(edge,15);

weightMetrics(row,col)=Featuresweight(i);

weightMetrics(col,row)=Featuresweight(i);

end

% weightMetrics=maxmin(weightMetrics);

[X,Y]=meshgrid(0.5:1:15.5,0.5:1:15.5);

pcolor(X,Y,weightMetrics);

colorbar;

colormap(flipud(hot));

set(gca,'YTick',[1:1:15]);

set(gca,'XTick',[1:1:15]);

set(gca,'XTickLabel',{'F3','FZ','F4','FC3','FCZ','FC4','C3','CZ','C4','CP3','CPZ','CP4','P3','PZ','P4','null'})

set(gca,'YTickLabel',{'F3','FZ','F4','FC3','FCZ','FC4','C3','CZ','C4','CP3','CPZ','CP4','P3','PZ','P4','null'})

% shading flat;

% axis off;

result=true;