median module¶

A module providing facilities for creating and saving median images.

A median image for a date is created by collating all the images taken that night and finding the median greyscale value for every pixel position. A method is provided to save the median image. Two other methods are used to support the median creation: one implementing the median of medains algorithm and one that converts a list of lists to a list of tuples.

median.median_all_date(date, camera='KPNO', color=False)[source]¶

Find the median images for a given date.

Parameters:	date (str) – The date to find median images for. camera ({"KPNO", "SW"}) – The camera used to take the image. “KPNO” represents the all-sky camera at Kitt-Peak. “SW” represents the spacewatch all-sky camera. color (bool, optional) – If true, finds the median images in color, otherwise works in grayscale. Defaults to False.
Returns:	A dictionary mapping exposure times to their median images.
Return type:	dict of ndarrays

See also

io_util.load_all_date(): Load images in color to find color medians.

median.median_of_medians(arr, i)[source]¶

Find the ith smallest element of a list using the median of medians algorithm.

Parameters:	arr (array_like) – An array_like object of floats. i (int) – The positional rank.
Returns:	The ith smallest element of arr.
Return type:	float

Notes

i = len(arr) // 2 corresponds to finding the median of arr. Details on the median of medians algorithm can be found at Wikipedia (https://en.wikipedia.org/wiki/Median_of_medians).

median.ndarray_to_tuplelist(arr)[source]¶

Convert an ndarray to a list of tuples.

For an ndarray of shape (y, x) returns a list of y tuples where each tuple is of length x.

Parameters:	arr (ndarray) – A NumPy ndarray to be converted.
Returns:	A list of tuples.
Return type:	list

median.save_medians(medians, date, color=False)[source]¶

Save a dict of medians produced by median_all_date.

Parameters:	medians (dict) – Dictionary mapping exposure times to their median images. date (str) – The date of the median images. color (bool, optional) – If True, saves the median images in color, otherwise works in grayscale. Defaults to False.