Quickstart

To use Gepard it is essential to understand the two main code objects:

  1. DataPoint contains information about kinematics and, possibly, about particular measurement perfomed at that kinematics

  2. Theory, contains algorithms for evaluation of various structure functions (CFFs, GPDs, …) and observables (cross-sections, asymmetries, …) for a given data point,

These objects are classes (in the sense of object-oriented programming) and one actually works with particular instances of these objects. If you don’t have experience with object-oriented programming paradigm, don’t worry, examples below explain everything that you need to know.

DataPoint

Instance of DataPoint can be constructed by passing kinematics information as Python keyword (named) arguments:

>>> import gepard as g
>>> pt = g.DataPoint(xB=0.348, t=-0.3, Q2=3., phi=0.3)

Information about some experimental measurement, performed at a given kinematic point, can be added:

>>> pt = g.DataPoint(xB=0.348, t=-0.3, Q2=3., phi=0.3,
...                   process='ep2epgamma', exptype='fixed target',
...                   in1energy=6., in1charge=-1, in1polarization=+1,
...                   observable='XS', val=0.21, err=0.01)

This particular datapoint corresponds to a measurement of DVCS, i. e., \(e p \to e p \gamma\), in a fixed target setting, where beam particle (target particle has attributes starting as in2...) has energy of 6 GeV2, negative charge (electron) and positive helicity. What is measured (observable) is cross-section (XS) and result of the measurement is 0.21 nb, with total uncertainty of 0.01 nb.

All attributes of datapoint are documented here.

This information can then be accessed as attributes of DataPoint object, using . (dot) operator:

>>> pt.xB
0.348
>>> pt.xi
0.210653753

where one notices that other, dependent kinematic variables are automatically precalulated, like \(\xi = x_B / (2 - x_B)\) here.

Datapoints can be organized in datasets (class DataSet), and, for convenience, many datasets, corresponding to various measurements performed over the years, are already made available within Gepard, as documented in section Working with datasets.

Theory

For convenience, several Theory objects are immediately available to the user who just wants to calculate observables. For example, KM15 model can be imported like this

>>> from gepard.fits import th_KM15

and then used to calculate theory prediction for a given datapoint

>>> th_KM15.predict(pt)
0.02344956

Method predict of the Theory object will by default calculate observable specified in observable attribute of pt. User can also calculate other observables, like beam charge asymmetry

>>> th_KM15.AC(pt)
0.138049

All implemented observables are listed here.

Furthermore, values of Compton Form Factors are available, for example \(\mathfrak{Im}\mathcal{H}\)

>>> th_KM15.ImH(pt)
2.80754427

Note

Presently, you cannot calculate observable or form factor by directly specifying kinematics, like



>>> # This will NOT work
>>> th_KM15.ImH(x=0.348, t=-0.3, Q2=3)

You have to create a DataPoint object first:

>>> pt = g.DataPoint(xB=0.348, t=-0.3, Q2=3)
>>> th_KM15.ImH(pt)  # This will work
2.8075