nucleon_elastic_ff.data.arraymanip

Routines for array manipulations

Routines are mostly an interface for numpy or cupy.

nucleon_elastic_ff.data.arraymanip.average_arrays(arrays: List[numpy.ndarray], axis: int = 0) → numpy.ndarray

Averages arrays over specified dimension. Input can be a list.

arrays: List[np.ndarray]

The arrays to average.

axis: int = 0

The average dimension index.

nucleon_elastic_ff.data.arraymanip.get_fft(array: numpy.ndarray, axes: Tuple[int] = (1, 2, 3), cuda: bool = False) → numpy.ndarray

Execute fft for input array over specified axes.

For input $f(t, n_i)$, the transformation is defined by $$ f(t, k_i) = \left( \prod_{i=1}^3 \sum_{n_i=0}^{N_i} \exp \left\{ - \frac{2 \pi i}{N_i} n_i k_i \right \} \right) f(t, n_i) $$

Arguments

mean: np.ndarray

The source averaged input data of shape [NT, NZ, NY, NX]

axes: Tuple[int] = (1, 2, 3)

The axes on which the fft is executed.

cuda: bool = False

Use cupy to do fft transformation.

nucleon_elastic_ff.data.arraymanip.shift_array(array: numpy.ndarray, shift: int = 0, axis: int = 0) → numpy.ndarray

Rolls the array in specified dimension by shift: v[n] -> v[n+shift]

array: List[np.ndarray]

The arrays to average.

shift: int

The amount to shift.

axis: int = 0

The shift dimension index.