match_fraction#

Methods for computing match fractions.

hsr4hci.match_fraction.get_all_match_fractions(residuals, roi_mask, hypotheses, parang, psf_template, frame_size, n_roi_splits=1, roi_split=0)[source]#

This is essentially a convenience function which wraps the loop over the ROI and calls get_match_fraction_for_position() for every spatial pixel.

Parameters:

  • residuals (Dict[str, ndarray]) – A dictionary containing the full residuals as they are produced by hsr4hci.training.train_all_models().

  • hypotheses (ndarray) – A 2D numpy array containing the hypotheses map.

  • parang (ndarray) – A 1D numpy array of shape (n_frames, ) containing the parallactic angle for every frame.

  • psf_template (ndarray) – A 2D numpy array containing the unsaturated PSF template.

  • frame_size (Tuple[int, int]) – A tuple (x_size, y_size) containing the spatial size of the input stack in pixels.

  • n_roi_splits (int) – Total number of splits for the ROI if we want to compute the match fraction map in parallel.

  • roi_split (int) – Index of the ROI split that we want to process here.

  • roi_mask (ndarray) –

Returns:

A 3-tuple consisting of

  1. mean_mfs: A 2D numpy array containing the match fraction map when using the mean to average.

  2. median_mfs: A 2D numpy array containing the match fraction map when using the median to average.

  3. affected_pixels: A 4D numpy array containing which, for each position (x, y) contains a 2D binary mask with the affected mask (see get_match_fraction_for_position()).

Return type:

Tuple[ndarray, ndarray, ndarray]

hsr4hci.match_fraction.get_match_fraction_for_position(position, hypothesis, residuals, parang, psf_template, signal_times, frame_size)[source]#

Compute the match fraction for a single given position.

Parameters:

  • position (Tuple[int, int]) – A tuple (x, y) specifying the position for which to compute the match fraction.

  • hypothesis (float) – The hypothesis (= temporal index) for the given position. In general, this should be an integer, but the type here has to be a float because the value may also be NaN (in case there is no hypothesis).

  • residuals (Dict[str, ndarray]) – A dictionary containing the full residuals as they are produced by hsr4hci.training.train_all_models().

  • parang (ndarray) – A 1D numpy array of shape (n_frames, ) containing the parallactic angle for every frame.

  • psf_template (ndarray) – A 2D numpy array containing the unsaturated PSF template.

  • signal_times (ndarray) – A 1D numpy array of shape (n_signal_times, ) containing the temporal grid.

  • frame_size (Tuple[int, int]) – A tuple (x_size, y_size) containing the spatial size of the input stack in pixels.

Returns:

A 3-tuple consisting of

  1. match_fraction__mean: The match fraction for the given target position when using the mean to average.

  2. match_fraction__median: The match fraction for the given target position when using the median to average.

  3. affected_mask: A 2D numpy array containing a binary mask that indicates the pixels from which the match fraction was computed (i.e., the pixels that are affected by the planet according to the hypothesis).

Return type:

Tuple[float, float, ndarray]