以对数刻度绘制(Matlab)

本文介绍了以对数刻度绘制(Matlab)的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！ 问题描述

我有一组数据，应该以对数标度标绘. x值始终为正，而y值为正和负.因此，loglog命令将只忽略负y值.但我想绘制它们:我想以对数标度表示y轴，但要同时具有负值和正值.

I have a set of data, which should be plotted in a loglog scale. x-values are always positive, but y-values are positive and negative. So, loglog command will just omit negative y-values. But I want to plot them: I want to have y-axis in logarithmic scale, but with negative values, as well as positive values.

基本上，如果我们有一组数据(x,y)， 我想绘制:(log(x),log(y))，如果y>0，和(log(x),-log(-y))，如果y<0.我尝试使用这些公式，然后使用简单的plot函数，但是使用这种方法时，轴不在对数范围内.再一次，我希望两个轴都为对数刻度.

Basically, if we have set of data (x,y), I want to plot: (log(x),log(y)), if y>0, and (log(x),-log(-y)), if y<0. I tried to to use these formula, and then simple plot function, but with this approach axes are not in logarithmic scale. Once again, I want both axes to be in logarithmic scale.

谢谢

为明确起见，我希望结果与plot(x,y)完全相同，但y轴和x轴均以对数标度

to clarify, I want the result to be exactly as plot(x,y), but both y-axis and x-axis to be in logarithmic scale

推荐答案

有一些用于绘制对数图的函数:沿y，沿x或两者皆有: semilogy ， semilogx ， loglog .您需要loglog一个:

There's a few functions for plotting logarithmic plots: along y, along x or both: semilogy, semilogx, loglog. You'll need the loglog one:

ypos = y(y>0); % Gets positive values xpos = x(y>0); % Get corresponding x values yneg = y(y<0); % Gets negative values xneg = x(y<0); % Get corresponding x values figure; loglog(xpos,ypos) hold on loglog(xneg,-yneg, 'r')

这基本上在同一图中创建了两个图，一个具有正y值，另一个具有负y值.简而言之:

This basically creates two plots in the same figure, one with the positive and one with the negative y values. In short:

figure; loglog(x(y>0),y(y>0)) hold on loglog(x(y<0),-y(y<0),'r')

由于按定义对数严格为正，因此您不能创建负标度.您可以做的就是通过设置来赋予您幻觉

Since the logarithm is by definition strict positive, you cannot create a negative scale. What you can do is give the illusion you have one by setting

set(gca, 'xdir','reverse')

这确实意味着您需要两个单独的图，以防止x轴也从正数向后移动.

This does mean you'd need two separate plots though, to prevent the x-axis from the positive numbers running backwards as well.