seberg

New ways to index NumPy arrays

Author

Sebastian Berg

Contact

sebastian@sipsolutions.net

Date

2015-07-14

Status

draft

seberg

import numpy as np

def object_einsum(string, *arrays):
    """Simplified object einsum, not as much error checking
    
    does not support "..." or list input and will see "...", etc. as three times
    an axes identifier, tries normal einsum first!
    
    NOTE: This is untested, and not fast, but object type is
    never really fast anyway...

seberg

import numpy as np
from numpy import asarray, add, rollaxis, sort, arange

def percentile(a, q, limit=None, interpolation='linear', axis=None,
               out=None, overwrite_input=False):
    """
    Compute the qth percentile of the data along the specified axis.

    Returns the qth percentile of the array elements.

seberg

def rolling_window(array, window=(0,), asteps=None, wsteps=None, axes=None, toend=True):
    """Create a view of `array` which for every point gives the n-dimensional
    neighbourhood of size window. New dimensions are added at the end of
    `array` or after the corresponding original dimension.
    
    Parameters
    ----------
    array : array_like
        Array to which the rolling window is applied.
    window : int or tuple

seberg

"""Impressive little thing how einsum + stride_tricks can beat numpys build in C functions
for correlate (for large data). (ok depending on the implementation of np.correlate, the
comparison is not fair, but still rather impressive that you can get comparable speeds)
"""

import numpy as np
import stride_tricks as st # stride_tricks.py gist

a = np.random.random((100,100,100)).ravel()
stamp = np.random.random((3,3,3)).ravel()

seberg

import numpy as np
from collections import defaultdict as _dd

def rolling_window(array, window=(0,), asteps=None, wsteps=None, axes=None, toend=True):
    """Create a view of `array` which for every point gives the n-dimensional
    neighbourhood of size window. New dimensions are added at the end of
    `array` or after the corresponding original dimension.
    
    Parameters
    ----------