pymchelper.estimator module

class pymchelper.estimator.ErrorEstimate

Bases: enum.IntEnum

When averaging data multiple files we could estimate statistical error of scored quantity. Such error can be calculated as: none (error information missing), standard error or standard deviation.

none = 0

stddev = 2

stderr = 1

class pymchelper.estimator.Estimator

Bases: object

Estimator data including scoring mesh description.

This class handles in universal way data generated with MC code. It includes data (data and data_raw fields) and optional errors (error and error_raw). Estimator holds also up to 3 binning axis (x, y and z fields). Scored quantity can be assigned a name (i.e. dose) and unit (i.e. Gy). Several other fields are also used:

• nstat: number of simulated histories
• counter: number of files read to construct detector object
• corename: common core part of input files defining a name of detector
• error_type: none, stderr or stddev - error type

Estimator data can be either read from the file (see fromfile method in input_output module or constructed directly:

>>> d = Estimator()
>>> d.x = MeshAxis(n=2, min_val=0.0, max_val=10.0, name="X", unit="cm", binning=MeshAxis.BinningType.linear)
>>> d.x.data
array([ 2.5,  7.5])
>>> d.y = MeshAxis(n=3, min_val=0.0, max_val=150.0, name="Y", unit="cm", binning=MeshAxis.BinningType.linear)
>>> d.y.data
array([  25.,   75.,  125.])
>>> d.z = MeshAxis(n=1, min_val=0.0, max_val=1.0, name="Z", unit="cm", binning=MeshAxis.BinningType.linear)
>>> d.z.data
array([ 0.5])

add_page(page)

Add a page to the estimator object. New copy of page is made and page estimator pointer is set to the estimator object holding this page. :param page: :return: None

axis(axis_id)

Mesh axis selector method based on integer id’s.

Instead of getting mesh axis data by calling d.x, d.y or d.z (assuming d an object of Detector class) we can get that data by calling d.axis(0), d.axis(1) or d.axis(2). See for example: >>> d = Estimator() >>> d.x = MeshAxis(n=2, min_val=0.0, max_val=10.0, name=”X”, unit=”cm”, binning=MeshAxis.BinningType.linear) >>> d.y = MeshAxis(n=3, min_val=0.0, max_val=150.0, name=”Y”, unit=”cm”, binning=MeshAxis.BinningType.linear) >>> d.z = MeshAxis(n=1, min_val=0.0, max_val=1.0, name=”Z”, unit=”cm”, binning=MeshAxis.BinningType.linear) >>> d.axis(1) MeshAxis(n=3, min_val=0.0, max_val=150.0, name=’Y’, unit=’cm’, binning=<BinningType.linear: 0>)

Parameters:axis_id – axis id (0, 1 or 2) Returns:MeshAxis object

dimension

Let’s take again detector d with YZ scoring. >>> e = Estimator() >>> e.x = MeshAxis(n=1, min_val=0.0, max_val=1.0, name=”X”, unit=”cm”, binning=MeshAxis.BinningType.linear) >>> e.y = MeshAxis(n=3, min_val=0.0, max_val=150.0, name=”Y”, unit=”cm”, binning=MeshAxis.BinningType.linear) >>> e.z = MeshAxis(n=2, min_val=0.0, max_val=2.0, name=”Z”, unit=”cm”, binning=MeshAxis.BinningType.linear) >>> e.dimension 2

Returns:number of axes (among X,Y,Z) which have more than one bin

pymchelper.estimator.average_with_nan(estimator_list, error_estimate=<ErrorEstimate.stderr: 1>)

Calculate average estimator object, excluding malformed data (NaN) from averaging. :param estimator_list: :param error_estimate: :return: