bayes : Fit simulated profiles against experiment

class idpflex.bayes.TabulatedFunctionModel(xdata, ydata, interpolator_kind='linear', prefix='', missing=None, name=None, **kwargs)[source]

Bases: lmfit.model.Model

A fit model that uses a table of (x, y) values to interpolate

Uses interp1d

Fitting parameters:
  • integrated intensity amplitude \(A\)
  • position of the peak center \(E_0\)
  • nominal relaxation time tau \(\tau\)
  • stretching exponent beta \(\beta\)
Parameters:
  • xdata (ndarray) – X-values to construct the interpolator
  • ydata (ndarray) – Y-values to construct the interpolator
  • interpolator_kind (str) – Interpolator that interp1d should use
guess(y, x=None, **kwargs)[source]

Estimate fitting parameters from input data

Parameters:
  • y (ndarray) – Values to fit to, e.g., SANS or SAXS intensity values
  • x (ndarray) – independent variable, e.g., momentum transfer
Returns:

Parameters with estimated initial values.
Return type:

Parameters
idpflex.bayes.fit_at_depth(tree, experiment, property_name, depth)[source]

Fit at a particular tree depth from the root node

Fit experiment against the property stored in the nodes. The fit model is generated by model_at_depth()

Parameters:
  • tree (Tree) – Hierarchical tree
  • experiment (ProfileProperty) – A property containing the experimental info.
  • property_name (str) – The name of the simulated property to compare against experiment
  • max_depth (int) – Fit at each depth up to (and including) max_depth
Returns:

Results of the fit
Return type:

ModelResult
idpflex.bayes.fit_to_depth(tree, experiment, property_name, max_depth=5)[source]

Fit at each tree depth from the root node up to a maximum depth

Fit experiment against the property stored in the nodes. The fit model is generated by model_at_depth()

Parameters:
  • tree (Tree) – Hierarchical tree
  • experiment (ProfileProperty) – A property containing the experimental info.
  • property_name (str) – The name of the simulated property to compare against experiment
  • max_depth (int) – Fit at each depth up to (and including) max_depth
Returns:

A list of ModelResult items containing the fit at each level of the tree up to and including max_depth
Return type:

list
idpflex.bayes.model_at_depth(tree, depth, property_name)[source]

Generate a fit model at a particular tree depth

Parameters:
  • tree (Tree) – Hierarchical tree
  • depth (int) – depth level, starting from the tree’s root (depth=0)
  • property_name (str) – Name of the property to create the model for
Returns:

A model composed of a TabulatedFunctionModel for each node plus a ConstantModel accounting for a flat background
Return type:

CompositeModel
idpflex.bayes.model_at_node(node, property_name)[source]

Generate fit model as a tabulated function with a scaling parameter, plus a flat background

Parameters:
  • node (ClusterNodeX) – One node of the hierarchical Tree
  • property_name (str) – Name of the property to create the model for
Returns:

A model composed of a TabulatedFunctionModel and a ConstantModel
Return type:

CompositeModel