以预定义的宽度/高度比将matplotlib图保存在右侧的图例中 [英] Save matplotlib plot with legend on the right in predefined width/height ratio

问题描述

我想将一个 matplotlib 图保存为一个宽/高比为 1.25 的 png 文件.我通过figsize参数指定了该比率.但是，当我使用选项 bbox_inches ="tight" 保存图形时，输出png的大小为553至396像素，比率为1.39.我想保留 bbox_inches = "tight" 选项以防止图形边框中出现不必要的空白.我尝试了类似 stackoverflow 帖子中建议的不同方法，但找不到解决方案.

这是示例代码:

 将matplotlib.pyplot导入为plt将numpy导入为npfig = plt.figure(figsize = (3, 2.4), dpi = 150)ax = plt.subplot(111)对于范围内的 i (3):ax.plot(np.random.random(10)，np.random.random(10)，"o"，label = i)ax.legend(bbox_to_anchor =(1，0.6)，title ="Title")plt.ylabel("标签")plt.xlabel("标签")plt.title("标题", loc = "left")plt.savefig("test.png", format = "png", dpi = 150, bbox_inches = "tight")

这是输出的png

解决方案

bbox_inches = "tight" 明确地告诉 matplotlib 裁剪或扩展图形；因此，图形大小的任何设置都将丢失.因此，如果要控制图形尺寸，则不能使用此选项.

您还有其他选择:

定义BBox

• 定义您自己的 bbox_inches，它确实具有所需的方面.
  该图像现在为547 x 438像素，因此纵横比为1.2488，与您可以达到的1.25尽可能接近.

调整内边距

• 使用原始图形尺寸(3，2.4)并调整填充，以使所有元素都适合图形.这将使用 fig.subplots_adjust() 完成.

  fig = plt.figure(figsize = (3, 2.4), dpi = 150)fig.subplots_adjust(top=0.89,底部=0.195，左= 0.21，右 = 0.76)

  
  此图像现在具有 (3, 2.4)*150 = 450 x 360 像素的预期大小.

• 对于子图参数的自动确定，也看看这个问题:这个答案到如何将图例从情节中排除".

  I want to save a matplotlib figure as a png file with a width/height ratio of 1.25. I specified this ratio via the figsize argument. But when I save the figure using the option bbox_inches = "tight" then the output png has a size of 553 to 396 pixels which is a ratio of 1.39. I would like to keep the bbox_inches = "tight" option to prevent unnecessary white space in the figure borders. I tried different approaches suggested in similar posts on stackoverflow but couldn't figure out a solution.

  Here is example code:

  import matplotlib.pyplot as plt
  import numpy as np
  
  fig = plt.figure(figsize = (3, 2.4), dpi = 150)
  ax = plt.subplot(111)
  
  for i in range(3):
      ax.plot(np.random.random(10), np.random.random(10), "o", label = i)
  ax.legend(bbox_to_anchor=(1, 0.6), title = "Title")
  plt.ylabel("Label")
  plt.xlabel("Label")
  plt.title("Title", loc = "left")
  plt.savefig("test.png", format = "png", dpi = 150, bbox_inches = "tight")
  

  This is the output png

  解决方案

  The bbox_inches = "tight" explicitely tells matplotlib to crop or expand the figure; any settings for the figure size will hence be lost. Thus you cannot use this option if you want to have control over the figure size.

  Other options you have:

  Define BBox

  • Define your own bbox_inches, which does have the desired aspect. The dimensions of the Bbox would be [[x0,y0],[x1,y1]].

    import matplotlib.transforms
    bbox = matplotlib.transforms.Bbox([[-0.2, -0.36], [3.45, 2.56]])
    plt.savefig("test.png", format = "png", dpi = 150,bbox_inches =bbox)
    

    
    This image is now 547 x 438 pixels, thus having an aspect of 1.2488, which is as close as you can get to 1.25.

  Adjust padding

  • Use the original figure size of (3, 2.4) and adjust the padding, such all elements fit into the figure. This would be done using fig.subplots_adjust().

    fig = plt.figure(figsize = (3, 2.4), dpi = 150)
    fig.subplots_adjust(top=0.89,
                        bottom=0.195,
                        left=0.21,
                        right=0.76)
    

    
    This image now has the expected size of (3, 2.4)*150 = 450 x 360 pixels.

  • For an automatic determination of the subplot parameters, also look at this question: Creating figure with exact size and no padding (and legend outside the axes)

  In general, I would recommend reading this answer to "How to put the legend out of the plot`.

  这篇关于以预定义的宽度/高度比将matplotlib图保存在右侧的图例中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IT屋！