使用matlab绘画柱状图，且使用不同的图案填充

机器学习



在论文中，图表往往发挥着极为重要的作用，好的图表将能进一步提升论文的质量。在书写论文时，很多时候需要绘制柱状图，然而不同的柱状图如果采用颜色区分，当论文打印以后，视觉效果大打折扣，甚至无法区分。在遇到这个问题时，我通过网站论坛搜索，终于找到了在matlab中绘制柱状图，并采用不同的图案进行表示。主要利用下面的代码。

代码出自：http://www.aos.wisc.edu/~dvimont/matlab/Graphics_Tools/applyhatch.html

function applyhatch(h,patterns,colorlist)
%APPLYHATCH Apply hatched patterns to a figure
%  APPLYHATCH(H,PATTERNS) creates a new figure from the figure H by
%  replacing distinct colors in H with the black and white
%  patterns in PATTERNS. The format for PATTERNS can be
%    a string of the characters ‘/’, ‘\’, ‘|’, ‘-‘, ‘+’, ‘x’, ‘.’
%    a cell array of matrices of zeros (white) and ones (black)
%
%  APPLYHATCH(H,PATTERNS,COLORS) maps the colors in the n by 3
%  matrix COLORS to PATTERNS. Each row of COLORS specifies an RGB
%  color value.
%
%  Note this function makes a bitmap image of H and so is limited
%  to low-resolution, bitmap output.
%
%  Example 1:
%    bar(rand(3,4));
%    applyhatch(gcf,’\-x.’);
%
%  Example 2:
%    colormap(cool(6));
%    pie(rand(6,1));
%    legend(‘Jan’,’Feb’,’Mar’,’Apr’,’May’,’Jun’);
%    applyhatch(gcf,’|-+.\/’,cool(6));
%
%  See also: MAKEHATCH

%  By Ben Hinkle, [email protected]
%  This code is in the public domain.
 
oldppmode = get(h,’paperpositionmode’);
oldunits = get(h,’units’);
set(h,’paperpositionmode’,’auto’);
set(h,’units’,’pixels’);
figsize = get(h,’position’);
if nargin == 2
  colorlist = [];
end
bits = hardcopy(h,’-dzbuffer’,’-r0′);
set(h,’paperpositionmode’,oldppmode);

bwidth = size(bits,2);
bheight = size(bits,1);
bsize = bwidth * bheight;
if ~isempty(colorlist)
  colorlist = uint8(255*colorlist);
  [colors,colori] = nextnonbw(0,colorlist,bits);
else
  colors = (bits(:,:,1) ~= bits(:,:,2)) | …
           (bits(:,:,1) ~= bits(:,:,3));
end
pati = 1;
colorind = find(colors);
while ~isempty(colorind)
  colorval(1) = bits(colorind(1));
  colorval(2) = bits(colorind(1)+bsize);
  colorval(3) = bits(colorind(1)+2*bsize);
  if iscell(patterns)
    pattern = patterns{pati};
  elseif isa(patterns,’char’)
    pattern = makehatch(patterns(pati));
  else
    pattern = patterns;
  end
  pattern = uint8(255*(1-pattern));
  pheight = size(pattern,2);
  pwidth = size(pattern,1);
  ratioh = ceil(bheight/pheight);
  ratiow = ceil(bwidth/pwidth);
  bigpattern = repmat(pattern,[ratioh ratiow]);
  if ratioh*pheight > bheight
    bigpattern(bheight+1:end,:) = [];
  end
  if ratiow*pwidth > bwidth
    bigpattern(:,bwidth+1:end) = [];
  end
  bigpattern = repmat(bigpattern,[1 1 3]);
  color = (bits(:,:,1) == colorval(1)) & …
          (bits(:,:,2) == colorval(2)) & …
          (bits(:,:,3) == colorval(3));
  color = repmat(color,[1 1 3]);
  bits(color) = bigpattern(color);
  if ~isempty(colorlist)
    [colors,colori] = nextnonbw(colori,colorlist,bits);
  else
    colors = (bits(:,:,1) ~= bits(:,:,2)) | …
             (bits(:,:,1) ~= bits(:,:,3));
  end
  colorind = find(colors);
  pati = (pati + 1);
  if pati > length(patterns)
    pati = 1;
  end
end

newfig = figure(‘units’,’pixels’,’visible’,’off’);
imaxes = axes(‘parent’,newfig,’units’,’pixels’);
im = image(bits,’parent’,imaxes);
fpos = get(newfig,’position’);
set(newfig,’position’,[fpos(1:2) figsize(3) figsize(4)+1]);
set(imaxes,’position’,[0 0 figsize(3) figsize(4)+1],’visible’,’off’);
set(newfig,’visible’,’on’);

function [colors,out] = nextnonbw(ind,colorlist,bits)
out = ind+1;
colors = [];
while out <= size(colorlist,1)
  if isequal(colorlist(out,:),[255 255 255]) | …
        isequal(colorlist(out,:),[0 0 0])
    out = out+1;
  else
    colors = (colorlist(out,1) == bits(:,:,1)) & …
             (colorlist(out,2) == bits(:,:,2)) & …
             (colorlist(out,3) == bits(:,:,3));
    return
  end
end

而applyhatch函数需要调用下面的函数

function A = makehatch(hatch)
%MAKEHATCH Predefined hatch patterns
%  MAKEHATCH(HATCH) returns a matrix with the hatch pattern for HATCH
%   according to the following table:
%      HATCH        pattern
%     ——-      ———
%        /          right-slanted lines
%        \          left-slanted lines
%        |          vertical lines
%        –          horizontal lines
%        +          crossing vertical and horizontal lines
%        x          criss-crossing lines
%        .          single dots
%
%  See also: APPLYHATCH

%  By Ben Hinkle, [email protected]
%  This code is in the public domain.

n = 6;
A=zeros(n);
switch (hatch)
case ‘/’
  A = fliplr(eye(n));
case ‘\’
  A = eye(n);
case ‘|’
  A(:,1) = 1;
case ‘-‘
  A(1,:) = 1;
case ‘+’
  A(:,1) = 1;
  A(1,:) = 1;
case ‘x’
  A = eye(n) | fliplr(diag(ones(n-1,1),-1));
case ‘.’
  A(1:2,1:2)=1;
otherwise
  error([‘Undefined hatch pattern “‘ hatch ‘”.’]);
end

效果如下图所示

 

example:

data = [96.3,92.6,71.2;95.7,93.6,83.9;96.8,94.3,78.3;95.8,92.7,80.3]
bar(data,1)
axis([0 6 0.0 100])
legend(‘方法’,’exited’,’Square’)
set(gca,’XTickLabel’,{‘Img1′,’Img2′,’Img3′,’Img4’})
applyhatch(gcf,’\.x.’)

如果要只显示两个柱图:

程序如下更改：

data = [0.63694,0.68476;0,0;]
bar(data,1)
legend(‘WE\_DBA’,’Proposed Scheme”)
set(gca,’XTickLabel’,{‘WE_DBA’,’Proposed Scheme’})
applyhatch(gcf,’\.x.’)

上述的函数文件在程序中直接使用即可