Question

我正在尝试使用以下代码将使用hist2d生成的多个2D直方图保存到使用PdfPages生成的多页pdf中：

import numpy as np
import pandas as pd
import seaborn as sns
import scipy.stats as stats
import matplotlib.pyplot as plt
from matplotlib.backends.backend_pdf import PdfPages
import warnings 
import subprocess
import os
warnings.simplefilter("ignore", category=PendingDeprecationWarning)

x1 = np.random.randn(100000)
y1 = np.random.randn(100000) + 5

pp = PdfPages("somepdf.pdf")
fig = plt.figure()
plt.hist2d(x=x1,y=y2, bins=50)
plt.title(row['smRNAname'])
plt.xlabel("Position(BP)")
plt.ylabel("Read Length")
cb = plt.colorbar()
cb.set_label('counts in bin')
pp.savefig(fig, dpi=300, transparent = True)
plt.close()

fig = plt.figure()
fig = plt.hist2d(x=x1,y=y1, bins=50)
plt.title(row['PIWIname'])
plt.xlabel("Position(BP)")
plt.ylabel("Read Length")
cb = plt.colorbar()
cb.set_label('counts in bin')
pp.savefig(fig, dpi=300, transparent = True)
plt.close()
pp.close()

但是我遇到以下错误：

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-87-ccbb61958687> in <module>()
     61     cb = plt.colorbar()
     62     cb.set_label('counts in bin')
---> 63     pp.savefig(fig, dpi=300, transparent = True)
     64     plt.close()
     65 

/anaconda3/lib/python3.7/site-packages/matplotlib/backends/backend_pdf.py in savefig(self, figure, **kwargs)
   2519                 manager = Gcf.get_active()
   2520             else:
-> 2521                 manager = Gcf.get_fig_manager(figure)
   2522             if manager is None:
   2523                 raise ValueError("No figure {}".format(figure))

/anaconda3/lib/python3.7/site-packages/matplotlib/_pylab_helpers.py in get_fig_manager(cls, num)
     39         figure and return the manager; otherwise return *None*.
     40         """
---> 41         manager = cls.figs.get(num, None)
     42         if manager is not None:
     43             cls.set_active(manager)

TypeError: unhashable type: 'numpy.ndarray'

从错误本身来看，我可以理解，这可能是由于hist2d返回2D数组而不是引用fig（？）。直接使用plt.savefig（“ test.pdf”）保存直方图就可以了。我不确定自己在做什么错，还是不可能？

Answer 1

我相信问题出在此

fig = plt.figure()
fig = plt.hist2d(x=x1,y=y1, bins=50) <---- here should be plt.hist2d(x=x1,y=y1, bins=50)
plt.title(row['PIWIname'])

让我知道它是否有效：）

Python：无法使用PdfPages将matplotlib hist2d图保存到文件

1 个答案: