IMP::test Namespace Reference
[Python]

---

Detailed Description

Methods and classes for testing the IMP kernel and modules.

Data Structures
class	ApplicationTestCase
	Super class for simple IMP application test cases. More...
class	ConstPairScore
	An IMP::PairScore which always returns a constant value. More...
class	ConstQuadScore
	An IMP::PairScore which always returns a constant value. More...
class	ConstSingletonScore
	An IMP::SingletonScore which always returns a constant value. More...
class	ConstTripletScore
	An IMP::TripletScore which always returns a constant value. More...
class	ConstUnaryFunction
	An IMP::UnaryFunction which always returns a constant value. More...
class	DirectorObjectChecker
	Check to make sure the number of director references is as expected. More...
class	LogPairScore
	An IMP::PairScore which logs which Particles it is called with. More...
class	RefCountChecker
	Check to make sure the number of C++ object references is as expected. More...
class	TestCase
	Super class for IMP test cases. More...
class	TestRefiner
	A class which makes sure that the right particles are passed back. More...
Functions
def	numerical_derivative
	Calculate the derivative of the single-value function `func` at point `val`.
def	xyz_numerical_derivatives
	Calculate the x,y and z derivatives of `model`'s scoring function on the `xyz` particle.
Variables
	_basis_vector
	_model
	_starting_coordinates
	_xyz

---

Function Documentation

def IMP::test::numerical_derivative	(		func,
			val,
			step
	)

Calculate the derivative of the single-value function `func` at point `val`.

The derivative is calculated using simple finite differences starting with the given `step`; Richardson extrapolation is then used to extrapolate the derivative at step=0.

def IMP::test::xyz_numerical_derivatives	(		model,
			xyz,
			step
	)

Calculate the x,y and z derivatives of `model`'s scoring function on the `xyz` particle.

The derivatives are approximated numerically using the numerical_derivatives() function.