本文介绍了为什么此代码中需要numpy.ravel()才能产生较小的倍数?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

我找到了一些代码来生成一组小倍数,并且运行良好.

  fig,axes = plt.subplots(6,3,figsize =(21,21))fig.subplots_adjust(hspace=.3, wspace=.175)对于 ax,zip 中的数据(axes.ravel(), clean_sets):ax.plot(data.ETo,o")

for ax, data in zip(axes.ravel(), clean_sets): contians .ravel() 但我不明白这实际上在做什么或为什么有必要.

如果我看看文档我发现了以下内容:

返回连续的扁平化数组.

返回包含输入元素的一维数组.仅在需要时进行复制.

我猜对应于来自 plt.subplot() 的轴的返回是一个无法迭代的多维数组,但我真的不确定.一个简单的解释将不胜感激.


在这种情况下使用 .ravel() 的目的是什么?

解决方案

您的猜测是正确的. plt.subplots()返回一个由多个轴组成的 Axes numpy 数组,具体取决于输入.如果用参数 nrows ncols 定义2D网格,则返回的 numpy 数组也将是2D数组.

此行为在 pyplot.subplots squeeze 参数中的文档

squeeze :布尔,可选,默认值:True
如果为True,则从返回的Axes对象中挤出额外的尺寸:

  • 如果仅构造了一个子图(nrows = ncols = 1),则返回的单个Axes对象将作为标量返回.

  • 对于 Nx1 或 1xN 子图,返回的对象是一个 1D numpy 对象数组,Axes 对象作为 numpy 1D 数组返回.
  • 对于 NxM,N>1 和 M>1 的子图作为二维数组返回.

如果为 False,则根本不进行压缩:返回的 Axes 对象始终是包含 Axes 实例的二维数组,即使它最终是 1x1.

因为这里有 plt.subplots(6,3),因此 N>1, M>1,结果对象必然是一个 2D 数组,独立于 squeeze 设置为什么.

这使得有必要展平此数组,以便能够对其进行 zip 压缩.选项是

  • zip(axes.ravel())
  • zip(axes.flatten())
  • zip(axes.flat)

I found some code to generate a set of small multiples and it is working perfectly.

fig, axes = plt.subplots(6,3, figsize=(21,21))
fig.subplots_adjust(hspace=.3, wspace=.175)
for ax, data in zip(axes.ravel(), clean_sets):
    ax.plot(data.ETo, "o")

The line for ax, data in zip(axes.ravel(), clean_sets): contians .ravel() but I do not understand what this is actually doing or why it is necessary.

If I take a look at the docs I find the following:

I guess the return that corresponds to axes from plt.subplot() is a multidimensional array that can't be iterated over, but really I'm not sure. A simple explanation would be greatly appreciated.


What is the purpose of using .ravel() in this case?

解决方案

Your guess is correct. plt.subplots() returns either an Axes or a numpy array of several axes, depending on the input. In case a 2D grid is defined by the arguments nrows and ncols, the returned numpy array will be a 2D array as well.

This behaviour is explained in the pyplot.subplots documentation inside the squeeze argument,

  • for Nx1 or 1xN subplots, the returned object is a 1D numpy object array of Axes objects are returned as numpy 1D arrays.
  • for NxM, subplots with N>1 and M>1 are returned as a 2D arrays.

Since here you have plt.subplots(6,3) and hence N>1, M>1, the resulting object is necessarily a 2D array, independent of what squeeze is set to.

This makes it necessary to flatten this array in order to be able to zip it. Options are

  • zip(axes.ravel())
  • zip(axes.flatten())
  • zip(axes.flat)

这篇关于为什么此代码中需要numpy.ravel()才能产生较小的倍数?的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持!

09-26 23:34