amep.utils.weighted_runningmean2d

amep.utils.weighted_runningmean2d(data: ndarray, nav: int, kernel: str | ndarray = 'homogenous', mode: str = 'valid', width: float = 1.0) → ndarray

Compute the running mean of two-dimensional data.

This is of a 2d density field.

Parameters:

• data (np.ndarray) – Input data of shape (N1,N2,).

• nav (int) – length of the kernel if standard is chosen

• kernel (str|np.ndarray) – Kind of kernel to be used for weights. Possible are homogenous, triangle and gauss. Provided a fitting array will also use normed version of this to weight. Weight array should have positive entries. Otherwise use np.convolve directly.

• mode (str) – ‘same’, ‘valid’, or ‘full’

• width (float) – width of the gaussian. Only effective when using gaussian kernel.

Returns:

Averaged data.

Return type:

np.ndarray

Examples

>>> import amep
>>> import numpy as np
>>> a = np.zeros((7,7))
>>> a[3,3] = 1
>>> print(a)
[[0. 0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 1. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0. 0.]]
>>> print(amep.utils.weighted_runningmean2d(a, 3, mode='valid'))
[[0.         0.         0.         0.         0.        ]
 [0.         0.11111111 0.11111111 0.11111111 0.        ]
 [0.         0.11111111 0.11111111 0.11111111 0.        ]
 [0.         0.11111111 0.11111111 0.11111111 0.        ]
 [0.         0.         0.         0.         0.        ]]
>>>