Response

class ompy.Response(path)

Bases: object

Interpolates response read from file for current setup

Implementaion of following method Guttormsen et al., NIM A 374 (1996) 371–376. DOI:10.1016/0168-9002(96)00197-0

Throughout the class, be aware that “compton” mat refer to all non-discrete structures (instead of the real Compton effect only).

Variables:

• resp (pd.DataFrame) – Information of the response table

• compton_matrix (np.ndarray) – array with compton counts. Shape is (N_incident, N_cmp).

• Ecmp_array (np.ndarray) – energy array for the compton counts

• smooth_compton (bool) – If True, the compoton array is smoothed before further processing. defaults to False

• truncate (float) – After how many sigma to truncate gaussian smoothing. Defaults to 6.

The resonse object is initialized with the path to the source files required to perform the interpolation. The path varaiable can either be a folder (assuming “old” format) or a zip file containing the files otherwise found in the folder in the “old” format.

Parameters:

path (str or Path) – Path to the required file(s)

Todo

• adapt rutines for the possibility that not all cmp spectra have the same binning

Methods Summary

E_compton(Eg, theta)	Calculates the energy of an electron that is scattered an angle theta by a gamma-ray of energy Eg.
LoadDir(path[, resp_name, spec_prefix])	Method for loading response file and compton spectra from a folder.
LoadZip(path[, resp_name, spec_prefix])	Method for loading response file and compton spectra from zipfile.
dE_dtheta(Eg, theta)	Function to correct number of counts due to delta(theta) Adapted from MAMA in the file kelvin.f It is dE/dtheta of the E(theta) in Eq.
discrete_peaks(i_response, fwhm_abs_array)	Add discrete peaks for a given channel and smooth them
fan_method_compton(E, compton, i_start, i_stop)	Fan method
fan_to_end(E, compton, i_start, i_stop, ...)	Linear(!) fan interpolation from Compton edge to Emax
get_closest_compton(E)	Find and rebin closest energies from available response functions
get_probabilities()	Interpolate full-energy peak probabilities (...)
interpolate([Eout, fwhm_abs, return_table])	Interpolated the response matrix
iterpolate_checks()	Check on the inputs to interpolate
linear_backscatter(E, compton)	Interpolate one-to-one up to the backscatter peak
linear_cmp_interpolation(E, compton[, ...])	Linear interpolation between the compton spectra
linear_to_end(E, compton, i_start, i_stop)	Interpolate one-to-one from the last fan energy to the Emax
two_channel_split(E_centroid, E_array)	When E_centroid is between two bins in E_array, this function returns the indices of the two nearest bins and the distance to the lower bin.

Methods Documentation

static E_compton(Eg, theta)

Calculates the energy of an electron that is scattered an angle theta by a gamma-ray of energy Eg.

Note

For Eg <= 0.1 it returns Eg. (workaround)

Parameters:

• Eg – Energy of incident gamma-ray in keV

• theta – Angle of scatter in radians

Returns:

Energy Ee of scattered electron

LoadDir(path, resp_name='resp.dat', spec_prefix='cmp')

Method for loading response file and compton spectra from a folder.

Parameters:

• path (Union[str, Path]) – path to folder

• resp_name (Optional[str], optional) – name of file with response table

• spec_prefix (Optional[str], optional) – Prefix for all spectra

Returns:

(tuple) containing

• resp (DataFrame): Information of the response table.

• compton_matrix (ndarray): matrix with compton counts.

  Shape is (N_incident, N_cmp)

• last.E (ndarray): energy array

LoadZip(path, resp_name='resp.csv', spec_prefix='cmp')

Method for loading response file and compton spectra from zipfile.

Is assumes that path is a zip file that constains at least XX files. At least one has to be a special summary table to be named resp_name.

Parameters:

• path (Union[str, Path]) – path to folder

• resp_name (Optional[str], optional) – name of file with response table

• spec_prefix (Optional[str], optional) – Prefix for all spectra

Returns:

(tuple) containing

• resp (DataFrame): Information of the response table.

• compton_matrix (ndarray): matrix with compton counts.

  Shape is (N_incident, N_cmp)

• last.E (ndarray): energy array

static dE_dtheta(Eg, theta)

Function to correct number of counts due to delta(theta) Adapted from MAMA in the file kelvin.f It is dE/dtheta of the E(theta) in Eq. (2) in Guttormsen 1996.

Parameters:

• Eg – Energy of gamma-ray in keV

• theta – Angle of scatter in radians

Returns:

dE_dtheta

Return type:

TYPE

discrete_peaks(i_response, fwhm_abs_array)

Add discrete peaks for a given channel and smooth them

Parameters:

• i_response (int) – Channel in response matrix

• fwhm_abs_array (np.ndarray) – Array with fwhms for each channel

Return type:

ndarray

Returns:

Array with smoothed discrete peaks

fan_method_compton(E, compton, i_start, i_stop)

Fan method

Parameters:

• E (float) – Incident energy

• compton (dict) – Dict. with information about the compton spectra to interpolate between

• i_start (int) – Index where to start (usually end of backscatter)

• i_stop (int) – Index where to stop (usually E+n*resolution). Note that it can be stopped earlier, which will be reported through i_last

Returns:

R is Response for E, and i_last last

index of fan-method

Return type:

Tuple[np.ndarray, int]

fan_to_end(E, compton, i_start, i_stop, fwhm_abs_array)

Linear(!) fan interpolation from Compton edge to Emax

The fan-part is “scaled” by the distance between the Compton edge and max(E). To get a reasonable scaling, we have to use ~6 sigma.

Note

We extrapolate up to self.N_out, and not i_stop, as a workaround connected to Magne’s 1/10th FWHM unfolding [which results in a very small i_stop.]

Parameters:

• E (float) – Incident energy

• compton (dict) – Dict. with information about the compton spectra to interpolate between

• i_start (int) – Index where to start (usually end of fan method)

• i_stop (int) – Index where to stop (usually E+n*resolution)

• fwhm_abs_array (np.ndarray) – FHWM array, absolute values

Returns:

Response for E

Return type:

np.ndarray

get_closest_compton(E)

Find and rebin closest energies from available response functions

If E < self.resp[‘Eg’].min() the compton matrix will be replaced by an array of zeros.

Parameters:

E (float) – Description

Return type:

Dict[str, Any]

Returns:

Dict with entries Elow and Ehigh, and ilow and ihigh, the (indices) of closest energies. The arrays counts_low and counts_high are the corresponding arrays of compton spectra.

get_probabilities()

Interpolate full-energy peak probabilities (…)

interpolate(Eout=None, fwhm_abs=None, return_table=False)

Interpolated the response matrix

Perform the interpolation for the energy range specified in Eout with FWHM at 1332 keV given by FWHM_abs (in keV).

The interpolation is split into energy regions. Below the back-scattering energy Ebsc we interpolate linearly, then we apply the “fan method” (Guttormsen 1996) in the region from Ebsc up to the Compton edge, with a Compton scattering angle dependent interpolation. From the Compton edge to Egmax we also use a fan, but with a linear interpolation.

Note

Below the ~350 keV we only use a linear interpolation, as the fan method does not work. This is not described in Guttormsen 1996.

Parameters:

• folderpath – The path to the folder containing Compton spectra and

• resp.dat –

• Eout_array – The desired energies of the output response matrix.

• fwhm_abs (Optional[float]) – The experimental absolute full-width-half-max at 1.33 MeV. Note: In the article it is recommended to use 1/10 of the real FWHM for unfolding.

• return_table (optional) – Returns “all” output, see below

Returns:

• response (Matrix): Response matrix with incident energy on the “Ex” axis and the spectral response on the “Eg” axis

• response_table (DataFrame, optional): Table with efficiencies, FE, SE (…) probabilities, and so on

Return type:

Matrix or Tuple[Matrix, pd.DataFrame]

iterpolate_checks()

Check on the inputs to interpolate

linear_backscatter(E, compton)

Interpolate one-to-one up to the backscatter peak

Parameters:

• E (float) – Incident energy

• compton (dict) – Dict. with information about the compton spectra to interpolate between

Returns:

R is Response for E, and

i_bc is index of backscatter peak

Return type:

Tuple[np.ndarray, int]

linear_cmp_interpolation(E, compton, fill_value='extrapolate')

Linear interpolation between the compton spectra

Parameters:

• E (float) – Incident energy

• compton (dict) – Dict. with information about the compton spectra to interpolate between

• fill_value (str, optional) – Fill value beyond boundaries

Returns:

Interpolated values

Return type:

f_cmp (nd.array)

linear_to_end(E, compton, i_start, i_stop)

Interpolate one-to-one from the last fan energy to the Emax

Parameters:

• E (float) – Incident energy

• compton (dict) – Dict. with information about the compton spectra to interpolate between

• i_start (int) – Index where to start (usually end of fan method)

• i_stop (int) – Index where to stop (usually E+n*resolution)

Returns:

Response for E

Return type:

np.ndarray

static two_channel_split(E_centroid, E_array)

When E_centroid is between two bins in E_array, this function returns the indices of the two nearest bins and the distance to the lower bin. The distance to the higher bin is 1-floor_distance

Parameters:

• E_centroid (double) – The energy of the centroid (mid-bin)

• E_array (np.array, double) – The energy grid to distribute