How to initialize a matrix to a very large number, say to infinity.
Similar to initializing all elements to zero: sample = np.matrix((np.zeros(50,50))
I want to initalize to infinity
How to do it in python?
Asked
Active
Viewed 2.0k times
5
-
You must understand that a computer exists in a finite world. Perhaps what you're asking is how to dynamically allocate an array. – Lescurel Sep 11 '17 at 08:42
-
"infinity" is _not_ "a very large number". – tobias_k Sep 11 '17 at 08:44
-
2@Lescurel infinity *is* a value representable in float numbers. Check out [the format specification for IEEE 754](https://en.wikipedia.org/wiki/IEEE_754-1985) for details on how that works – GPhilo Sep 11 '17 at 08:45
-
1@GPhilo Thanks, I've learned something today – Lescurel Sep 11 '17 at 08:54
2 Answers
12
There is np.infin numpy
>>> sample = np.matrix(np.ones((50,50)) * np.inf)
>>> sample
matrix([[ inf, inf, inf, ..., inf, inf, inf],
[ inf, inf, inf, ..., inf, inf, inf],
[ inf, inf, inf, ..., inf, inf, inf],
...,
[ inf, inf, inf, ..., inf, inf, inf],
[ inf, inf, inf, ..., inf, inf, inf],
[ inf, inf, inf, ..., inf, inf, inf]])
Ghilas BELHADJ
- 13,412
- 10
- 59
- 99
-
Use of this, import numpy as np sample = np.matrix(np.ones(4,4) * np.inf) shows an error a = empty(shape, dtype, order) TypeError: data type not understood – Santle Camilus Sep 11 '17 at 09:11
-
@Dr.SantleCamilus You missed a couple of brackets inside `np.ones()` – Ghilas BELHADJ Sep 11 '17 at 09:13
-
-
is there a way I can mix random integers with 2d matrix having infinity values? – May 14 '21 at 06:36
