Finding nice color bar limits automatically

Question

William Thielicke 2024-1-21

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2072471-finding-nice-color-bar-limits-automatically

评论： David Goodmanson 2024-1-22

Dear all, I have a plot with a colorbar. The limits of the color bar can't be adjusted automatically, because the plot is a bit complicated with RGB conversion etc. So I want to apply my own automatic upper and lower limit to the color bar. These limits should "look nice": if the data ranges e.g. from -0.06535 to 7.6682, the limits should be something like -0.1 and 8 (or 7.9...?). If the data ranges from 21 to 493 it should range from 20 to 500 or so. If the data ranges from 0.00653 to 0.0954 the limits should e.g. be 0.005 and 0.01. You probably noticed that I have difficulties to define what "looks nice", do you have a suggestion for my problem? Thanks a lot!

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

David Goodmanson 2024-1-21

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2072471-finding-nice-color-bar-limits-automatically#answer_1393641

编辑：David Goodmanson 2024-1-21

在 MATLAB Online 中打开

Hi William,

It looks like you want to move from the limits you have to some with fewer significant figures such as 1 or 2. Here is one possible answer with some functions to try.

function y = roundn(x,n)
% round x to n significant figures.
% rounds toward nearest integer in last significant figure.
% y = roundn(x,n)
logflr = floor(log10(abs(x)));
pof10 = 10.^(n-1-logflr);
y = round(x.*pof10)./pof10;

function y = fixn(x,n)
% round x to n significant figures.
% rounds towards zero.
% y = fixn(x,n)
logflr = floor(log10(abs(x)));
pof10 = 10.^(n-1-logflr);
y = fix(x.*pof10)./pof10;

function y = floorn(x,n)
% round x to n significant figures.
% rounds toward minus inf.
% y = floorn(x,n)
logflr = floor(log10(abs(x)));
pof10 = 10.^(n-1-logflr);
y = floor(x.*pof10)./pof10;

function y = ceiln(x,n)
% round x to n significant figures.
% rounds toward plus inf.
% y = ceiln(x,n)
logflr = floor(log10(abs(x)));
pof10 = 10.^(n-1-logflr);
y = ceil(x.*pof10)./pof10;