Running GenX Models in Jupyter Notebook/Lab

Introduction

Jupyter notebooks have become quite popular in experimental and data science as they allow to work on data analysis and write code interactively while being able to reproduce the results and display results nicely.

While the main strength of GenX is the interactive development of models using a graphical user interface while keeping the felxibility of a modifyable model script, GenX also supports to use the essential features via an API interface that includes special jupyter notebook support.

While it is possible to build models via this API it is mostly intended to run and anlyze already existing model, e.g. to perform statistical analysis of the results.

Examples

The genx installation includes some example notebooks that can be used as a reference. You can find them in:

{genx installation}/genx/examples/*.ipynb

Quick Start

To use the GenX API, the package needs to be found by your python environment use in the notebook. I you installed GenX via pip in the same environment, that should already be the case.

The following shows you how to load a GenX file, inspect the model and run a fit:

%pytlab inline # %matplotlib inline
from genx import api

model,optimizer=api.load('D17_TOF_SiO.hgx')
# save the model script to a file for later inspection
with open('genx_model.py', 'w') as fh:
    fh.write(model.script)

########## new cell ############
model.parameters
# ==> output a list of fit parameters from the "grid" tab of the GUI
########## new cell ############
optimizer
# ==> output the configuration options for the fit, can be modified as attributes of this object
########## new cell ############
api.fit_notebook(model, optimizer)
# ==> performs the fit and shows live update of the progress
########## new cell ############
api.fit_update(model, optimizer)
model
# ==> update model with fit results and show the overview widget for the model
########## new cell ############
# potentially save the fit result to a new file
#api.save(r'testoutput.hgx', model, optimizer)

Widgets

There are some experimental notebook widgets implemented, that allow interactive modification of GenX models similar to the GUI. This can be useful for testing or if only partial code based analysis is desired. The widgets can be accessed as attributes of some objects: