The solution result of numpy gradient function is inconsistent when the variables are array and matrix.

topic description

use the numpy gradient function to calculate the gradient

related codes

/ / Please paste the code text below (do not replace the code with pictures)

g = img [0 3, 0:3]

g =
array ([[138144141],

   [134, 135, 132],
   [137, 135, 133]], dtype=uint8)

gy, gx = np.gradient (g)
gy=
array ([252. , 247. , 247. ],

   [127.5, 123.5, 124. ],
   [  3. ,   0. ,   1. ]])

gx=
array ([6,1.5,253. ],

   [  1. , 127. , 253. ],
   [254. , 126. , 254. ]])

arr=np.array ([[138144141], [134135132], [137135133]])
arr=
array ([[138144141],

)

   [134, 135, 132],
   [137, 135, 133]])

gy, gx = np.gradient (arr)
gy=
array ([- 4. 9. 9. ],

   [-0.5, -4.5, -4. ],
   [ 3. ,  0. ,  1. ]])

gx=
array ([6. 1. 5,-3. ],

   [ 1. , -1. , -3. ],
   [-2. , -2. , -2. ]])

what result do you expect? What is the error message actually seen?

Why are the two results inconsistent? how is the gradient calculated under the matrix type?

Numpy python

Apr.05,2021

data overflow, convert the matrix type to float and then calculate the gradient

Previous: How does nginx determine whether a file has different access paths according to its existence?

Next: Problems with el-upload uploading files using axios

I have 30, 000 pictures now. How can I save them in numpy format?
as mentioned above, I am going to save about 30, 000 pictures in numpy format first, so that it will be much more convenient and faster to use later. but in the process of running, it seems that there is an error due to insufficient memory, which is r...

Gpu numpy python

Feb.28,2021
The python that can be executed normally in the interactive environment reports an error in IDE.
1. According to Google s official machine learning course code, there is no problem when executing in an interactive environment, but an error will be reported when executed in vscode; the 2.python code is as follows: import numpy as np np.random.p...

Visual-studio-code numpy python

Mar.01,2021
Random walk question ask why the argmax output will be a lot of 0? It didn't live up to expectations.
this is a basic random walk problem. according to the contents of the book, I try to find indexes on axis=1 whose absolute value is greater than or equal to 7, but there are always several rows whose index is 0, and the index entered normally is fine....

Numpy python

Mar.02,2021
Numpy index exercise, ask for advice
ask God to explain how the results of this a [iMagnej] are all taken out of a. Why do I not understand the meaning expressed by [iMagnej] ? ...

Numpy python

Mar.16,2021
Where do the properties contained in python come from?
for example, numpy, I looked at the package structure of numpy on github and found that numpy is a package, but its _ _ init__.py file does not have a function definition such as ones (), and there is no import reference to it, so why can I write numpy.o...

Numpy python

Mar.24,2021
The sum according to the condition in the matrix?
problem description The matrix is as follows: 1,2017,2,5 1,2018,1,5 1,2018,2,4 2,2017,1,5 1,2018,1,2 according to the first three columns, there are three kinds of matrix to produce a new matrix: 1,2017,2,5 1,2018,1,7 1,2018,2,4 2,2017,1,5 one li...

Numpy python

Mar.28,2021
Why is the calculation of Fibonacci by numpy matrix method 10 times faster than that of GMP built-in fib function?
didn t GMP say the fastest large number operation library on earth? ...

Numpy python

Aug.17,2021
Why do two identical list to np.array, data types become object, and the other float??
Why do two list-to-np.array, datatypes of exactly the same type become object, and the other floatable! print(len(X_l)) -sharp 2013 print(len(X_s)) -sharp 1980 print(X_l[0].dtype) -sharp float64 print(X_s[0].dtype) -sharp float64 print(X_l[0].shape) ...

Numpy python

Nov.08,2021
May I ask how to effectively speed up python multiple nested loops?
involves the calculation of a large amount of data there are multiple loops nested within the use of pandas for calculation extremely slow is there any effective way to speed up? ...

Numpy python pandas

Jul.19,2022