pyplot.subplots:python 和 jupyter notebook 中的不同行为 [英] pyplot.subplots: different behavior in python and jupyter notebook

问题描述

在参加 kaggle 比赛时，我遇到了一些奇怪的问题.基本上，我试图将图像的矢量表示转换为 png 文件.它在 iPython 中完美运行，代码如下:

While taking part in kaggle competition, i got some weird problem. Basically, I am trying to convert vector representation of am image to png file. It worked perfectly in iPython, code below:

def drawing_to_np_prepare_data(drawing):
    drawing = eval(drawing)
    fig, ax = plt.subplots()
    plt.close(fig)
    print('[debug] ax=',ax)
    for x,y in drawing:
        ax.plot(x, y, marker='.')
        ax.axis('off')
    fig.canvas.draw()

    # Convert images to numpy array
    np_drawing = np.array(fig.canvas.renderer._renderer)

    print('[debug] fig_size=',fig.get_size_inches())
    print('[debug] dpi=',fig.dpi)

    print('[debug] shape=',np_drawing.shape)
    print('[debug] size=',np_drawing.size)
    print('[debug] shape=',np_drawing.shape)

    im = cv2.cvtColor(np_drawing.astype(np.uint8), cv2.COLOR_BGR2RGB)

    # compress
    compressed_array = io.BytesIO()
    np.savez_compressed(compressed_array, im)
    compressed_array.seek(0)
    print('[debug] size=',np_drawing.shape)
    return compressed_array 

结果显示:

[debug] ax=AxesSubplot(0.125,0.125;0.775x0.755)
[debug] fig_size= [6. 4.]
[debug] dpi= 72.0
[debug] np_drawing.size= 497664
[debug] shape= (288, 432, 4)
[debug] size= 1880

可以满足我的需求:我得到的图像压缩后的尺寸<2Kb

which satisfy my needs: i am getting an image with compressed size < 2Kb

然而，当我从 CLI 在 python 中运行这段代码时，我得到了完全不同的结果:

However, when I run this code in python from CLI, I am getting quite different result:

[debug] ax=AxesSubplot(0.125,0.11;0.775x0.77)
[debug] fig_size= [6.4 4.8]
[debug] dpi= 100.0
[debug] np_drawing.size= 1228800
[debug] shape= (480, 640, 4)
[debug] size= 13096

如您所见，图形大小、dpi、轴不同，因此最后的大小也不同.

as you can see, figure size, dpi, axes are different and as a result, size at the end are also different.

我可以将参数传递给子图:

I can pass arguments to subplots:

plt.subplots(figsize=(6.,4.), dpi=72)

可校正除轴以外的参数(由于轴的不同，我猜是和尺寸):

which corrects parameters except axes (and size, I guess because of different axes):

[debug] ax=AxesSubplot(0.125,0.11;0.775x0.77)
[debug] fig_size= [6. 4.]
[debug] dpi= 72.0
[debug] np_drawing.size= 497664
[debug] shape= (288, 432, 4)
[debug] size= 8214

注意:我已经检查了库的版本，它们是相同的.

Note: I've checked library versions and they are the same.

因此，出现了多个问题:

So, multiple questions arise:

1. 为什么子图会给出不同的轴，形状和分辨率?

1. Why subplots give different axes, shape and resolution?

如何修正坐标轴?

如何在python中获得相同的行为?

How to get the same behaviors in python?

我想了解发生了什么.谢谢！

I want to understand the what is going on. Thanks!

推荐答案

要在脚本中获得与笔记本中完全相同的设置，请打开笔记本，运行

To get the exact same settings in a script as in your notebook, open a notebook, run

%matplotlib inline
%config InlineBackend.rc

它将打印一个 rcParams 字典.

It'll print a dictionary of rcParams.

{'figure.figsize': (6.0, 4.0),
 'figure.facecolor': (1, 1, 1, 0),
 'figure.edgecolor': (1, 1, 1, 0),
 'font.size': 10,
 'figure.dpi': 72,
 'figure.subplot.bottom': 0.125}

将这些内容复制为您的python文件

Copy those to your python file as

newrc = {'figure.figsize': (6.0, 4.0),
         'figure.facecolor': (1, 1, 1, 0),
         'figure.edgecolor': (1, 1, 1, 0),
         'font.size': 10,
         'figure.dpi': 72,
         'figure.subplot.bottom': 0.125}

import matplotlib.pyplot as plt
plt.rcParams.update(newrc)

然后做你的情节.

这是否真的解决了不同渲染器尺寸的问题无法测试，因为该问题不包含可运行示例.

Whether or not this actually solves the problem of different renderer sizes cannot be tested because the question does not contain a runnable example.

这篇关于pyplot.subplots:python 和 jupyter notebook 中的不同行为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！