-
Notifications
You must be signed in to change notification settings - Fork 0
/
redblue.m
118 lines (116 loc) · 4.15 KB
/
redblue.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
function y = redblue(varargin)
% generate a RED-BLUE colormap with zero as white.
% Syntax: y = redblue(n,clim,'black')
% Typical usage: colormap(redblue(64))
% Positive values are displayed as blue intensities and negative
% values displayed as red intensities. Zero is white.
% The clim values are used to find zero to set that to white.
% Arguments:
% All arguments are optional and can be in any order.
% n - number of color levels
% (optional with default as # of colors of current colormap)
% clim - two element vector specifying the color limits
% (optional with default as current axis color limits)
% black - string ('k' or 'black') specifying zero as black.
% (optional with default as zero is white)
% See: colormap
%
% Notes:
% This creates a custom colormap for any image so that the value
% zero is either white or black. The colorbar scale will be skewed
% toward red or blue depending on the caxis values of the image.
%
% This version flattens the edges of the colormap to improve the
% visualization of the gradient. This is better for larger values of n.
%
% Keep in mind that if the scale is very skewed, there will not
% be much of a color gradient. The gradient can always be increased
% by using your own clim values.
% Example:
% y = caxis; % e.g. y = [-11,-5]
% colormap(redblue(64) % and not much gradient
% colormap(redblue(64,[-11,0])) % white is -5 with larger gradient
% created 3/24/2020 Mirko Hrovat [email protected]
% modified 5/04/2020 Changed algorithm for colormap calculation, eliminating
% an error which occurs for small "clim" differences.
ntotmax = 1e4;
n = []; clim = []; black = false;
for k = 1:nargin
a = varargin{k};
switch true
case ischar(a)
switch true
case strcmpi(a,'k')
black = true;
case strcmpi(a,'black')
black = true;
end
case isnumeric(a) && numel(a) == 1
n = a;
case isnumeric(a) && numel(a) == 2
clim = a;
end
end
if isempty(n)
n = size(colormap,1);
end
if isempty(clim)
clim = caxis;
end
cmin = min(clim);
cmax = max(clim);
% np controls the visual flattening of extreme regions.
% It is the number of color levels used in these extreme regions.
% 1/8 of the range (0 to 1) uses a different gradient slope.
% Extreme regions are at max red, min red, max blue, min blue.
% np=n/16 provides a uniform gradient but is less appealing visually.
np = max(round(sqrt(n)/4),1); % scales slowly with n
switch true
case cmin >= 0 % display just blue
ntot = round(n*cmax/(cmax-cmin));
if ntot > ntotmax % check if cmax is close to cmin
ntot = ntotmax;
end
n2 = ntot - 2*np;
v = [linspace(1,.875,np),linspace(.875,.125,n2),...
linspace(.125,0,np)];
y = repmat(v',[1,3]);
y(:,3) = 1;
y(1:ntot-n,:) = [];
case cmax <= 0 % display just red
ntot = round(n*abs(cmin/(cmax-cmin)));
if ntot > ntotmax % check if cmax is close to cmin
ntot = ntotmax;
end
n2 = ntot - 2*np;
v = [linspace(0,.125,np),linspace(.125,.875,n2),...
linspace(.875,1,np)];
y = repmat(v',[1,3]);
y(:,1) = 1;
y(n+1:end,:) = [];
otherwise % display both red and blue
ntot = round(n*2*max(cmax,abs(cmin))/(cmax-cmin));
if ntot > ntotmax % check if cmax is close to cmin
ntot = ntotmax;
end
m = fix(ntot/2);
n2 = round((ntot-4*np)/2);
v = [linspace(-1,-.875,np),linspace(-.875,-.125,n2),...
linspace(-.125,.125,2*np),linspace(.125,.875,n2),...
linspace(.875,1,np)];
y = 1-abs(repmat(v',[1,3]));
y(end-m+1:end,3) = 1; % set blue
y(1:m,1) = 1; % set red
if cmax > abs(cmin)
y(1:ntot-n,:) = [];
else
y(n+1:end,:) = [];
end
end
if black
y2 = y;
y(:,1) = (1-y2(:,2)).*(y2(:,1)==1);
y(:,3) = (1-y2(:,2)).*(y2(:,3)==1);
y(:,2) = 0;
end
end