Structure Function API

class structurefunction.SFResult(med, err_low, err_high, count, c_bins)
c_bins: ndarray

Alias for field number 4

count: ndarray

Alias for field number 3

err_high: ndarray

Alias for field number 2

err_low: ndarray

Alias for field number 1

med: ndarray

Alias for field number 0

class structurefunction.TqdmToLogger(logger, level=None)[source]

Output stream for TQDM which will output to logger module instead of the StdOut.

flush()[source]

Flush write buffers, if applicable.

This is not implemented for read-only and non-blocking streams.

write(buf)[source]

Write string to file.

Returns the number of characters written, which is always equal to the length of the string.

structurefunction.bilby_fit(x: ~numpy.ndarray, y: ~numpy.ndarray, y_err: ~numpy.ndarray, outdir: str, label: str, model=<function broken_power_law>, **kwargs) → Result[source]

Bilby fit

Parameters:

  • x (np.ndarray) – X data

  • y (np.ndarray) – Y data

  • y_err (np.ndarray) – Y error

  • outdir (str) – Output directory

  • label (str) – Label

  • model (func, optional) – Model function. Defaults to broken_power_law.

Raises:

NotImplementedError – If model is not implemented

Returns:

Fitting result

Return type:

bilby.core.result.Result

structurefunction.broken_power_law(x: ndarray, amplitude: float, x_break: float, alpha_1: float, alpha_2: float) → ndarray[source]

Broken power law model

Parameters:

  • x (np.ndarray) – Frequency

  • amplitude (float) – Amplitude

  • x_break (float) – Break frequency

  • alpha_1 (float) – Power law index below break frequency

  • alpha_2 (float) – Power law index above break frequency

Returns:

Model array

Return type:

np.ndarray

structurefunction.combinate(data: ndarray) → Tuple[ndarray, ndarray][source]

Return all combinations of data with itself

Parameters:

data (np.ndarray) – Data to combine.

Returns:

Data_1 matched with Data_2

Return type:

Tuple[np.ndarray, np.ndarray]

structurefunction.fit_data(sf_result: SFResult, fit: str = 'bilby', outdir: str | None = None, model_name: str | None = None, n_point: int = 2, show_plots: bool = False, save_plots: bool = False, **kwargs) → None | Result[source]

Fit the structure function data

Parameters:

  • sf_result (SFResult) – Structure function result

  • fit (str, optional) – Fit type. Defaults to “bilby”.

  • outdir (str, optional) – Output directory for bilby. Defaults to None.

  • model_name (str, optional) – Model to fit. Defaults to None.

  • n_point (int, optional) – Number of points in SF. Defaults to 2.

  • show_plots (bool, optional) – Show fitting plots. Defaults to False.

  • save_plots (bool, optional) – Save fitting plots. Defaults to False.

Raises:

  • NotImplementedError – If model_name is not implemented

  • ValueError – If fit is not implemented.

Returns:

_description_

Return type:

Union[None, bilby.core.result.Result]

structurefunction.lsq_fit(x: ~numpy.ndarray, y: ~numpy.ndarray, outdir: str, label: str, model=<function broken_power_law>) → Result[source]

Least squares fit

Parameters:

  • x (np.ndarray) – X data

  • y (np.ndarray) – Y data

  • outdir (str) – Output directory

  • label (str) – Fitting label

  • model (func, optional) – Model function. Defaults to broken_power_law.

Raises:

NotImplementedError – if model is not implemented

Returns:

Fitting result

Return type:

Result

structurefunction.lsq_weight_fit(x: ~numpy.ndarray, y: ~numpy.ndarray, yerr: ~numpy.ndarray, outdir: str, label: str, model=<function broken_power_law>) → Result[source]

Weighted least squares fit

Parameters:

  • x (np.ndarray) – X data

  • y (np.ndarray) – Y data

  • yerr (np.ndarray) – Y error

  • outdir (str) – Output directory

  • label (str) – Label

  • model (func, optional) – Model function. Defaults to broken_power_law.

Raises:

NotImplementedError – If model is not implemented

Returns:

Fiting result

Return type:

bilby.core.result.Result

structurefunction.mc_sample(data: ndarray, errors: ndarray, samples: int = 1000) → ndarray[source]

Sample errors using Monte-Carlo Assuming Gaussian distribution.

Parameters:

  • data (np.ndarray) – Measurements

  • errors (np.ndarray) – Errors

  • samples (int, optional) – Samples of the distribution. Defaults to 1000.

Returns:

Sample array. Shape (len(data/errors),samples)

Return type:

np.ndarray

structurefunction.nanvar(data: ndarray | Quantity) → ndarray | Quantity[source]

Compute the variance of an array, ignoring NaNs

Parameters:

data (Union[np.ndarray, u.Quantity]) – Array

Returns:

Variance

Return type:

Union[np.ndarray, u.Quantity]

structurefunction.power_law(x: ndarray, amplitude: float, x_break: float, alpha: float) → ndarray[source]

Power law model

Parameters:

  • x (np.ndarray) – Frequency

  • amplitude (float) – Amplitude

  • x_break (float) – Reference frequency

  • alpha (float) – Power law index

Returns:

Model array

Return type:

np.ndarray

structurefunction.sf_three_point(rm_1: ndarray, rm_2: ndarray, rm_err_1: ndarray, rm_err_2: ndarray, src_1: ndarray, src_2: ndarray, dtheta: Quantity, bins: Quantity, bin_unit: Quantity) → SFResult[source]

Compute the three-point structure function

Parameters:

  • rm_1 (np.ndarray) – RMs of source 1 in pair (n_samples, n_pairs)

  • rm_2 (np.ndarray) – RMs of source 2 in pair (n_samples, n_pairs)

  • rm_err_1 (np.ndarray) – RM errors of source 1 in pair (n_samples, n_pairs)

  • rm_err_2 (np.ndarray) – RM errors of source 2 in pair (n_samples, n_pairs)

  • src_1 (np.ndarray) – Source 1 in pair (n_pairs)

  • src_2 (np.ndarray) – Source 2 in pair (n_pairs)

  • dtheta (u.Quantity) – Separation between sources in pair (n_pairs)

  • bins (u.Quantity) – Angular (or 3D Euclidean) separation bins

  • bin_unit (u.Quantity) – Unit to set bins to when resolving units

Returns:

Structure function result

Return type:

SFResult

structurefunction.sf_two_point(rm_1: ndarray, rm_2: ndarray, rm_err_1: ndarray, rm_err_2: ndarray, dtheta: Quantity, bins: Quantity, bin_unit: Quantity) → SFResult[source]

Compute the two-point structure function

Parameters:

  • rm_1 (np.ndarray) – RMs of source 1 in pair (n_samples, n_pairs)

  • rm_2 (np.ndarray) – RMs of source 2 in pair (n_samples, n_pairs)

  • rm_err_1 (np.ndarray) – RM errors of source 1 in pair (n_samples, n_pairs)

  • rm_err_2 (np.ndarray) – RM errors of source 2 in pair (n_samples, n_pairs)

  • dtheta (u.Quantity) – Separation between sources in pair (n_pairs)

  • bins (u.Quantity) – Angular (or 3D Euclidean) separation bins

  • bin_unit (u.Quantity) – Unit to set bins to when resolving units

Returns:

Structure function results

Return type:

SFResult

structurefunction.structure_function(data: Quantity, errors: Quantity, coords: SkyCoord, samples: int, bins: Quantity | int, show_plots: bool = False, save_plots: bool = False, verbose: bool = False, fit: str | None = None, outdir: str | None = None, model_name: str | None = None, n_point: int = 2, **kwargs) → Tuple[SFResult, Result][source]

Compute the second or third order structure function with Monte-Carlo error propagation.

Parameters:

  • data (u.Quantity) – 1D array of data values.

  • errors (u.Quantity) – 1D array of errors.

  • coords (SkyCoord) – 1D array of coordinates.

  • samples (int) – Number of samples to use for Monte-Carlo error propagation.

  • bins (Union[u.Quantity, int]) – Bin edges of the structure function, or number of bins.

  • show_plots (bool, optional) – Show plots. Defaults to False.

  • save_plots (bool, optional) – Save plots. Defaults to False.

  • verbose (bool, optional) – Print progress. Defaults to False.

  • fit (str, optional) – How to fit the broken powerlaw. Can be ‘astropy’, ‘astropy_mc’ or ‘bilby’. Defaults to None.

  • outdir (str, optional) – Output directory for bilby. Defaults to None.

  • model_name (str, optional) – Name of the model. Defaults to None. Can be ‘broken_power_law’ or ‘power_law’.

  • **kwargs – Additional keyword arguments to pass to the bilby.core.run_sampler function.

Returns:

The structure function result and the fitting result.

Return type:

Tuple[SFResult, bilby.core.result.Result]