调整地块大小并保存空

Question

我正在试图绘制一个热图重叠一个图像，使用海运和matplotlib，如下面的代码所示：

import seaborn as sns
import matplotlib
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import matplotlib.image as mpimg 

df = pd.read_csv('/home/diegonachon/eye-tracker/eyedata/csv_heatmap/cl/Suj5_Control_HO_1ph.csv')
df['xpl'] = pd.to_numeric(df['xpl'], errors = 'coerce')
df['ypl'] = pd.to_numeric(df['ypl'], errors = 'coerce')
df = df[(df['event'] == 'FIX') & (~(df['xpl'].isnull()))]
hmax = sns.kdeplot(x = df.xpl, y = df.ypl, cmap="Reds", shade=True, bw_adjust= 0.6, clip = ((0,800), (0,600)), 
                   alpha = 0.6, antialiased = True)
hmax.collections[0].set_alpha(0)

map_img = mpimg.imread('/home/diegonachon/eye-tracker/eyedata/asd.jpg') 

plt.imshow(map_img, zorder=0, extent=[0, 800, 0, 600])

# image = plt.imshow(map_img, zorder=0, extent=[0, 800, 0, 600], aspect = 250)
# figure(figsize=(8, 6), dpi=80)
# plt.imsave('test.png', image)
# plt.savefig('fig8_6_80dpi.png')   

plt.show()

结果：https://imgur.com/a/4rdgCOH

1. 第一件事是我无法调整它的大小，我尝试过更改“方面”和“图形大小”，但是没有发生任何事情(如注释中的代码所示)

当我试图保存绘图生成时，是和空文件(

1. )或错误(取消注释图像=plt.imswho().)- AttributeError：'tuple‘对象没有属性'shape')

谢谢!

Answer 1

下面是我对您的代码所做的更改。因为我不能访问你们的数据，所以我用了随机数，但是这会给你们一个想法。有关答案，请参阅大写注释。

import seaborn as sns
import matplotlib
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import matplotlib.image as mpimg 

# CREATED FOR TEST
x = np.random.rand(100,)
y = np.random.rand(100,)

fig, ax = plt.subplots(figsize=(10,10)) # YOU CAN MODIFY FIGSIZE FROM HERE
ax = sns.kdeplot(x = x, y = y, cmap="Reds", shade=True, bw_adjust= 0.6, 
clip = ((0,800), (0,600)), 
               alpha = 0.6, antialiased = True)
ax.collections[0].set_alpha(0)

# map_img = mpimg.imread('/home/diegonachon/eye-tracker/eyedata/asd.jpg') 

#plt.imshow(map_img, zorder=0, extent=[0, 800, 0, 600])

# image = plt.imshow(map_img, zorder=0, extent=[0, 800, 0, 600], 
#aspect = 250)
# figure(figsize=(8, 6), dpi=80)
# plt.imsave('test.png', image)
# plt.savefig('fig8_6_80dpi.png')  

# THIS WILL SAVE YOUR IMAGE AS A PNG FILE
plt.savefig('test.png') 
plt.show()