Question

我正在处理拟合图像，并尝试将IDL代码转换为python，但是我在理解如何转换IDL median()和rot()函数方面遇到困难。我有一个数组im_darksub，它是(ny, nx, n_im)，其中nx,ny是每个图像的形状，n_im是图像的数量（总共处理12个图像）。我试图翻译IDL代码im_median = median(im_darksub,dimension=1)和im_medrot = rot(im_median,85,cubic=-0.5)。我以为我可以将其翻译为np.median()和spicy.misc.imrotate，但是在使用imrotate时仍然会出错：

    Traceback (innermost last):
      File "<console>", line 1, in <module>
      File "/Users/courtneywatson1/Ureka/python/lib/python2.7/site-packages/scipy/misc/pilutil.py", line 345, in imrotate
        im = toimage(arr)
      File "/Users/courtneywatson1/Ureka/python/lib/python2.7/site-packages/scipy/misc/pilutil.py", line 234, in toimage
        raise ValueError("'arr' does not have a suitable array shape for "
    ValueError: 'arr' does not have a suitable array shape for any mode.

我猜是因为我的图像阵列超过2或3D？我使用IDL median()和rot()函数的方式是否等效于python？

Answer 1

这对我有用：

import numpy as np
x = np.arange(100)
x = np.reshape(x, (5, 5, 4))
im_median = np.median(x, axis=2)

import scipy.misc
im_rotate = scipy.misc.imrotate(y, 85, 'cubic')

我假设您看到imrotate已过时：

/Users/mgalloy/anaconda3/bin/ipython:1: DeprecationWarning: `imrotate` is deprecated!
`imrotate` is deprecated in SciPy 1.0.0, and will be removed in 1.2.0.
Use ``skimage.transform.rotate`` instead.

这似乎很好：

import skimage.transform
im_rotate = skimage.transform.rotate(im_median, 85, order=3, preserve_range=True)

Python相当于IDL中位数（）和rot（）

1 个答案: